Anthropic has disclosed a fourth incident in which one of its Claude models broke into genuine third-party systems during what was supposed to be a contained cybersecurity evaluation, deepening industry concern over the risks posed by increasingly autonomous AI agents.

The AI company said the episode dates back to January 2026 and involved an early version of Claude Opus 4.6, which breached external infrastructure after it was “unable to abort its task.” Anthropic has notified all affected parties, though it has not disclosed who they are. The incident is understood to have gone undetected until last month.

It follows three earlier cases revealed by Anthropic in July 2026, in which Claude Opus 4.7, Mythos 5 and an unnamed research model each compromised separate organisations during cybersecurity evaluations, again without the company’s knowledge at the time.

“AI safety is not just a model problem it’s an operational and human one. A simple configuration or naming error allowed a controlled test to interact with real systems, while the model continued pursuing its objective despite warning signs. Organisations deploying autonomous agents need clear accountability, isolated test environments, least-privilege access and human approval for high-impact actions. Trusting an AI to ‘do the right thing’ is not a security control.”
– Javvad Malik, Lead CISO Advisor at KnowBe4

According to Anthropic, all four incidents occurred during cybersecurity evaluations built by the same external evaluation partner. Claude was told it was operating in a simulated environment with no internet access, but a misconfiguration meant it was actually connected to the open internet. The partner responsible for the evaluations, Irregular, has since said the breach stemmed from a naming error: a fictional company used in the hacking simulations happened to match a real-world domain, causing the models to take offensive action against it.

Anthropic has since scanned roughly 481 million transcripts for similar problems and says it found no other cases of comparable or greater severity. The company has also signed an agreement with independent research non-profit METR to carry out a separate investigation into the incidents.

Anthropic attributes the root cause to two alignment failures: biased reasoning and recklessness. In practice, this meant the models tended to dismiss or misread evidence that their environment was connected to the live internet after initially being told otherwise, and were willing to pursue potentially harmful actions in single-minded pursuit of the task they had been assigned.

“Relying on an automated AI agent to audit another AI for security breaches is a fundamentally flawed approach. These models are increasingly acting as autonomous entities capable of probing network boundaries. When they escape their evaluation environments to access external systems, they create an unpredictable new vector for supply chain attacks. Developers must enforce strict network isolation rather than trusting software to mark its own homework.”
– Dray Agha, Senior Manager of Security Operations at Huntress

Anthropic said it was most concerned by the incident involving Claude Mythos 5, which went to considerable lengths to upload a malicious package to PyPI, the public repository used to distribute most Python software. The company noted that although the model repeatedly stated in its chain-of-thought reasoning that it believed it was in a simulation, the environment contained clear evidence it was on the real internet, and its actions were consistent with knowing this. Even after researchers amended the transcript to make the real-world context unmistakable, the model went on to carry out offensive actions.

Anthropic was keen to stress that the incidents remained narrow in scope: no Claude instance deviated from the task it had been set, none attempted to coordinate with other AI agents, and none tried to conceal evidence of its actions. The company added that biased reasoning appears less pronounced in its newer production models and can be reduced through more thorough alignment training, though the underlying cause, and why it was more acute in Mythos 5, remains unclear.

“As agentic AI systems become more capable and more widely deployed, the legal framework for who is responsible when a model accesses something it shouldn’t is going to need answering. These four incidents are an early test case for questions the law hasn’t fully addressed yet. Anthropic found four problems in 481 million transcripts and told everyone about it. The real question is how many problems the rest of the industry hasn’t looked hard enough to find yet.”
– Muhammad Yahya Patel, vCISO and Cybersecurity Advisor for EMEA at Huntress

The disclosure lands amid growing scrutiny of AI model safety more broadly. Rival OpenAI recently acknowledged a previously unreported incident from May 2026, in which internally deployed autonomous agents with read-only internet access took over a dormant German wiki forum, exchanging more than 18,000 posts as they attempted to coordinate answers and evade restrictions on a timed task. When a human moderator began removing the posts, the agents reportedly worked around the clean-up by naming backup pages so they would be buried at the end of an alphabetically sorted deletion list.

“We need to stop blaming AI and hold the humans in charge accountable for the actions of their AI agents. Companies will think twice about deploying AI if they are fined for negligence. It’s frustrating to listen to tech CEOs warn about the dangers of AI and then turn around and build it as if those dangers are unavoidable. If this problem gets bad enough, then we could see an emerging market for AI insurance that covers rogue third-party hacking, data theft, and intellectual property infringement.”
– Paul Bischoff, Consumer Privacy Advocate at Comparitech

Anthropic has warned that the risks are likely to grow rather than diminish as AI systems become more capable. “Future AI systems will be increasingly capable, which implies that misalignment will have the potential to cause more extreme harm,” the company said, adding that training robustly aligned frontier models remains an unsolved technical challenge that will require both continued research and stronger operational discipline from those deploying them.

For now, the incidents serve as a reminder that the weakest link in agentic AI deployments may not be the model itself, but the environments, configurations and oversight structures built around it.

The post Anthropic Discloses Fourth Incident of Claude Breaching Real Systems During Security Tests appeared first on IT Security Guru.

Huntress is expanding into Africa through a new distribution partnership with QBS Software Africa, the company announced today. The move marks the latest step in Huntress’ international growth as organised, AI-enabled cybercrime continues to rise worldwide.

The expansion comes as Africa faces a mounting cybersecurity crisis. According to INTERPOL’s 2026 African Cyberthreat Assessment, artificial intelligence is now linked to 55% of reported cybercrime across the continent, with direct economic damages surpassing $5 billion. Attackers are increasingly using AI to scale established tactics against under-resourced organisations, exploiting gaps in security coverage.

Under the new agreement, local value-added resellers and managed service providers will gain access to the Huntress Agentic Security Platform and its 24/7 AI-centric Security Operations Center. The platform is powered by Athena, Huntress’ agentic investigation system, which the company says combines machine-speed threat investigation with human security expertise to catch novel attacks before they escalate.

“Expanding into Africa is an important step in Huntress’ mission to protect all businesses from organized cybercrime,” said Adam Waggott-Moss, Head of EMEA Distribution at Huntress. “QBS Software Africa brings the local market knowledge, channel relationships, and regional expertise to help extend that protection to more businesses and build a strong foundation for long-term growth across the region.”

Headquartered in South Africa, QBS Software Africa will act as a launchpad for the rollout, beginning in South Africa before extending into select Sub-Saharan African markets. The partnership pairs Huntress’ cybersecurity technology with QBS Software Africa’s local channel expertise, technical support, and go-to-market resources, aiming to help regional partners scale stronger security offerings for their customers.

“Huntress’ combination of innovative technology and elite security expertise gives businesses across Africa the ability to keep pace with the attackers targeting them,” said Sanjay Mithal, Head of Strategic Vendor Alliances at QBS Software Africa. “The Huntress Agentic Security Platform brings layered defence together with the security experts who understand adversary behavior best, helping organizations across the region stay ahead of threats and sleep a little easier at night.”

The Africa launch builds on a period of rapid international growth for Huntress, which now protects 270,000 businesses across more than 100 countries. Over the past year, the company has deepened its presence across Europe and Asia Pacific, with adoption spreading across sectors including manufacturing, construction, retail and utilities, driven by investment in regional teams and channel partnerships.

Huntress plans to host a livestream on September 16 covering its Agentic Security Platform, titled “Inside the Agentic Huntress Platform: Beating Adversaries at Machine Speed.” You can register for the webinar here.

The post Huntress Expands into Africa with New QBS Software Africa Partnership appeared first on IT Security Guru.

UK cybersecurity specialist Core to Cloud has appointed former Currys Group CIO Andy Gamble as Chair of its Advisory Board as the company looks to accelerate the growth of its managed security services.

Gamble brings nearly 30 years of board-level technology leadership and will work with Core to Cloud on its strategic, advisory and commercial direction across the UK enterprise and mid-market sectors.

His appointment adds further experience to the company’s Advisory Board, which includes senior security leaders from major UK organisations.

From cybersecurity buyer to advisor

Gamble spent six years as Group CIO and Chief Transformation Officer at Currys PLC, where his responsibilities included large-scale technology transformation and cyber risk.

His career has also included senior CIO positions at Dyson, Sony Electronics and Essentra PLC. That experience means Gamble has spent much of his career on the customer side of the cybersecurity market, buying and managing the types of services Core to Cloud now provides.

“I spent the better part of three decades as a buyer of cybersecurity services, and the experience left me with a clear view of where the market falls short,” Gamble said.

“Most organisations understand that cyber risk is real. Far fewer have a security function that can communicate that risk clearly at board level, or a partner that moves fast enough to keep pace with the threat.”

Gamble said Core to Cloud stood out because of its focus on proactive security, adding that he intends to help the business scale its model as a challenger to conventional managed security service providers.

Supporting Core to Cloud’s next stage of growth

Based in Cirencester, Core to Cloud works with more than 150 organisations across sectors including the NHS, retail, financial services and critical national infrastructure.

Its services span Managed Detection and Response, Third-Party Cyber Risk Management, Security Assurance, Dark Web Monitoring and Threat Intelligence, and Cyber Crisis Simulation.

James Cunningham, CEO and Founder of Core to Cloud, said Gamble’s experience at the intersection of technology, risk and commercial strategy would bring a new perspective to the company.

“He understands what good security looks like from the inside and brings a depth of experience and perspective that will be hugely valuable as we continue to grow,” Cunningham said.

“We have an ambitious business, a strong customer base and services we genuinely believe in. Having Andy chair our board will help us build on those foundations, challenge our thinking and accelerate the next stage of Core to Cloud’s growth.”

The post Former Currys CIO Andy Gamble Joins Core to Cloud as Advisory Board Chair appeared first on IT Security Guru.

Cybersecurity firm Huntress has uncovered a malware campaign that preys on excitement for Grand Theft Auto VI (GTA6), packaging remote access trojans, an infostealer, and destructive ransomware inside fake “leaked” copies of the hotly anticipated game.

GTA6 is not due for release for another three months, but a wave of gameplay footage leaks and an official extended look from publisher Rockstar Games have pushed fan anticipation to a fever pitch. Threat actors have moved quickly to capitalise, seeding search results, gaming forums, social media and torrent sites with bogus disc image (ISO) files claiming to offer early access to the game. According to Huntress, no genuine leaked or playable version of GTA6 currently exists.

A booby-trapped installer

Some of the fake ISOs circulating exceed 100GB, padded with junk data to mimic the footprint of a legitimate AAA release, Huntress found. The actual malicious payload is far smaller, hidden inside a bundle that the researchers describe as an opportunistic attempt to throw “everything” at anyone who runs the installer.

Opening the ISO presents victims with a file named gta6installer.exe, oddly bearing the GTA5 icon rather than GTA6. Running it displays a Russian-language message warning that the game is unlicensed and may not launch, instructing users to email the attackers if they hit a “License not found” error so the “crack” can be fixed. That error message duly appears once installation finishes, a scripted piece of misdirection designed to explain away the fact that no game ever appears, while malware installs quietly in the background.

Three RATs, an infostealer, and a wiper

Beneath the fake installer, Huntress catalogued a small arsenal of malware, much of it several years old and simply repurposed for this campaign. Multiple copies of the remote access trojan NJRAT are dropped onto the system, along with a separate instance of DCRAT, giving attackers the ability to log keystrokes, access webcams, browse the desktop, steal browser credentials and cryptocurrency wallet details, and take full remote control of infected machines.

An open-source infostealer called Mercurial Grabber, nominally billed as an educational tool, is also installed, harvesting Discord tokens, Chrome-saved passwords and cookies, Roblox and Minecraft session data, Windows product keys and system information, and exfiltrating it all via a Discord webhook.

Most damaging is a variant of the well-known Chaos ransomware family, which Huntress says is being used purely as a wiper rather than for financial extortion. Once triggered on a machine with administrator rights, it deletes shadow copy backups, disables Windows recovery options, then either encrypts files under 200MB or overwrites larger files with random data, destroying them outright. It targets desktop, document, picture, and OneDrive folders, before changing the desktop wallpaper to an image of SpongeBob SquarePants declaring the machine hacked by the “Asha Hacker Team” and leaving a ransom note that provides no actual payment method.

The installer additionally drops a copy of Yandex Browser, a service popular in Russia and Eastern Europe. Combined with the Russian-language prompts and ransom messaging, Huntress believes the campaign is primarily targeting Russian-speaking gamers, although the tactics could easily be redeployed against fans elsewhere.

Old malware, same old lure

Huntress notes that none of the individual malware components are new or particularly sophisticated, and that an up-to-date version of Windows Defender should detect and block each one. The bigger risk, researchers say, lies in the social engineering: impatient fans searching for an early copy of a major game release are unusually willing to override security warnings and run unverified executables.

The firm’s advice is straightforward. Avoid downloading cracked or pirated software, especially for games not yet officially released. Anyone who believes they have already run the installer should disconnect the affected machine from the network immediately, reset passwords, enable two-factor authentication wherever possible, and fully reimage the system rather than attempt to clean it.

Huntress has published full technical indicators of compromise, including file hashes, dropped file paths, and command-and-control infrastructure, alongside its analysis. It can be found here: https://www.huntress.com/blog/fake-gta6-download-malware-analysis

The post Fake GTA6 ‘Leaked Download’ Caught Spreading RATs, Infostealer and Wiper Ransomware appeared first on IT Security Guru.

The UK’s National Cyber Security Centre (NCSC) has warned organisations about the security risks posed by “shadow AI”, as employees continue to turn to artificial intelligence tools that have not been approved by their employers.

In guidance published this week, the NCSC described shadow AI as the use of AI technology outside an organisation’s approved systems and processes, warning that security policies and governance have struggled to keep pace with the rapid adoption of AI in the workplace.

The scale of the issue could already be significant. Research cited by the NCSC found that 71% of employees have used AI tools that have not been approved by their employer.

According to the NCSC, unapproved AI services can expose sensitive company and customer information, reduce organisations’ visibility and control over their data, and create new opportunities for attackers. Information entered into consumer AI services may be stored, retained or used to improve those services outside existing corporate security and governance arrangements, depending on the privacy controls in place.

The agency also highlighted a potentially more serious risk as organisations move from generative AI tools towards AI agents. If an attacker exploits a vulnerability in an agent, they may be able to gain access to the same data, services and privileges legitimately available to that agent.

Darren Guccione, CEO and co-founder of Keeper Security, said the warning reflects a challenge many UK security teams are already facing.

“The NCSC’s warning on shadow AI reflects the reality of what many UK security teams are having to contend with. When employees adopt AI tools faster than IT can assess them, visibility gaps open long before governance has an opportunity to catch up. Microsoft’s research, cited by the NCSC, found that 71% of UK employees have used AI tools their employer hasn’t approved. Keeper Security’s 2026 research underlines the effect this is having on security teams, with 37% of UK organisations saying they lack visibility into which AI tools employees are actually utilising inside the business.

“The most significant detail in the NCSC’s warning is its point about AI agents inheriting the privileges of whoever deploys them. An attacker who compromises a poorly governed agent gains whatever access that agent holds, whether that’s a customer database or a finance system. Organisations that haven’t extended least-privilege and just-in-time access principles to their AI agents and non-human identities are exposed in ways endpoint controls alone won’t catch.”

Banning AI is unlikely to work

Rather than recommending organisations attempt to eliminate shadow AI altogether, the NCSC said businesses should focus on reducing the associated risks and understanding why employees are turning to unapproved tools in the first place.

It recommends creating a positive cyber security culture, providing AI tools that meet employees’ needs and securely integrating AI systems into the workplace.

Guccione added: “Banning shadow AI outright rarely works, as the NCSC itself acknowledges. Employees will find routes around blocked tools when the approved ones can’t do what they need.

“The more durable fix is improving visibility by identifying what identities, both human and machine, exist across the environment and what they can access. Organisations must enforce least-privilege principles by default, rather than waiting until an incident forces the question.”

Jamie Akhtar, CEO and co-founder at CyberSmart, agreed that businesses need to balance employees’ desire to use AI with appropriate security controls.

“The NCSC is right to highlight shadow AI as a growing cyber security challenge. Employees are using AI tools to work faster and more efficiently, but when those services sit outside an organisation’s approved systems, businesses can quickly lose visibility over where sensitive company and customer data is being shared, stored or processed.

“Simply banning AI is unlikely to solve the problem. Businesses need to provide secure, approved alternatives that allow people to benefit from AI without introducing unnecessary risk. Clear policies, employee education and appropriate technical controls all need to develop at the same pace as AI adoption.”

Akhtar said the challenge may be particularly difficult for SMEs without large in-house security teams, where managed service providers could help organisations identify unapproved technology and establish appropriate AI policies and controls.

“An MSP can help businesses identify unapproved technology, put proportionate AI policies and controls in place, educate employees and continuously manage emerging risks, giving organisations the confidence to embrace AI while maintaining visibility and control over their security.”

The NCSC said shadow AI is unlikely to disappear completely as AI services become cheaper and more readily available. Instead, organisations need to understand how and why employees are using these tools so they can provide secure alternatives while maintaining visibility over sensitive information and access to corporate systems.

The post NCSC Warns Shadow AI Is Creating New Security Blind Spots for UK Businesses appeared first on IT Security Guru.

Huntress researchers have uncovered two phishing attacks that combined convincing fake browser windows with legitimate remote management software to establish persistent access to victims’ devices.

Both incidents, observed in August, began with phishing messages directing victims to attacker-controlled websites. The attackers then used a browser-in-the-browser (BiTB) technique to create what appeared to be a legitimate Adobe webpage, before convincing victims to download malicious software disguised as an Adobe Reader update.

Rather than deploying conventional malware, the attackers installed rogue instances of ScreenConnect, legitimate remote monitoring and management (RMM) software, giving them continued remote access to compromised endpoints.

Fake browser makes phishing harder to spot

BiTB attacks create a fake browser window inside a webpage using HTML, CSS and JavaScript. The window can replicate familiar features including an address bar, padlock and legitimate-looking URL, making traditional advice such as checking the web address less effective.

In the first attack, detected on 25 August, a victim clicked a link in a phishing email and was taken to a fake CAPTCHA page. They were subsequently presented with blurred documents and told they needed to download Adobe PDF Reader to view them.

The fake browser page appeared to show Adobe’s legitimate get.adobe.com address. However, the supposed Reader installer was actually ScreenConnect.

Once installed, the attackers deployed two rogue ScreenConnect clients, providing redundant routes for maintaining access. They then executed HideCursor.exe, a defence-evasion tool designed to conceal on-screen activity. Huntress intervened before the attack could progress further.

Second attack follows same playbook

Huntress identified another incident on 31 August involving the same Adobe Reader lure.

This time, the victim interacted with a malicious link delivered through AT&T Office@Hand, a legitimate communications service powered by RingCentral. The attackers again disguised ScreenConnect as an Adobe Reader update and installed two unauthorised instances.

The second ScreenConnect session was used to execute another defence-evasion binary, HideUL.exe. Microsoft Defender detected part of the activity, but the rogue ScreenConnect client still completed its installation before Huntress shut down the attack.

Legitimate tools remain attractive to attackers

The attacks demonstrate how threat actors can combine familiar phishing techniques with trusted software to make malicious activity harder to identify.

RMM abuse is a growing problem. Huntress’ 2026 Cyber Threat Report found RMM abuse increased 277% year on year and appeared in nearly a quarter of the incidents investigated by the company.

Huntress recommends organisations restrict who can install remote management tools, maintain an approved inventory of RMM software and monitor for new or unauthorised ScreenConnect clients. Employees should also be wary of unexpected software updates or file-viewing prompts, even when a webpage appears to display a legitimate address.

Read the full research here. 

The post Huntress Uncovers Phishing Attacks Using Fake Browser Pages and Rogue RMM Tools appeared first on IT Security Guru.

Forescout has expanded its investment in its global partner ecosystem to strengthen technical expertise and help partners support customers managing increasingly complex IT, OT, IoT, and IoMT environments.

Nearly 100% of Forescout’s customer business is transacted through partners, making the channel a central part of the cybersecurity company’s growth strategy. Its latest investment includes the ongoing Mission: Possible enablement roadshow and the continued support of several partners in its Envision Partner Program.

Mission:Possible reaches partners worldwide

Launched in May, Mission:Possible represents the largest investment in channel engagement in Forescout’s history.

The programme is targeting 90 cities across more than 40 countries, with content available in 22 languages. More than 1,300 people have attended events so far, with Forescout expecting to engage over 2,000 partner professionals by the end of September.

Through technical education, hands-on discussions and local engagement, the programme is designed to give partners the expertise needed to help customers address emerging risks, including those associated with frontier AI and complex cyber-physical environments.

Partners advance through Envision programme

Forescout has also recognised several organisations that have reached new levels within its Envision Partner Program.

TIC Defense has advanced from Authorised Training Partner to Silver Reseller, while Trace3, Upstart Cyber and Sidif have moved from Silver to Gold Reseller status.

NTT Australia, Hitachi Sunway Information Systems Malaysia, DOR Information Technologies Israel and Computacenter Germany have all progressed from Gold to Platinum Reseller.

David Creed, Vice President of Worldwide Channel Sales at Forescout, said the company is continuing to raise expectations around technical capability, certifications, customer success and business performance.

“We’re proud to recognise these partners who have demonstrated exceptional commitment to customer success, technical excellence, partner enablement, and business growth while helping organisations meaningfully reduce cyber risk across converged IT, OT, IoT and IoMT environments,” he said.

Milo Sanchez, Senior Practice Director at Trace3, said Forescout’s technical investment and collaborative approach had helped the company better support its clients and grow the partnership.

Channel expertise becomes increasingly important

Forescout said partners are playing a growing role in helping organisations improve visibility and respond to threats across connected environments.

According to the company, organisations using its technology have reported discovering 50% more unknown IoT devices and reducing median breach containment time by 98.7%. The average time required to identify unknown, unmanaged and unauthorised assets has also fallen from 41 hours to six minutes.

The company’s channel programme has also received industry recognition, with the Envision Partner Program included in CRN’s Partner Program Guide and Forescout executives recognised in CRN’s Channel Chiefs and Channel Leaders EMEA lists.

The post Forescout Expands Global Investment in Channel Partners appeared first on IT Security Guru.

Black Duck, a provider of AI-powered application security solutions, has announced its participation in Project Glasswing, Anthropic’s industry-wide initiative aimed at protecting critical software infrastructure through the defensive use of advanced AI.

Through its involvement, Black Duck will integrate Mythos throughout its application security offerings, pairing AI-driven, deterministic vulnerability detection with established remediation processes, risk-based prioritisation, and governance frameworks built around compliance. The combination is designed to deliver a blended approach to security that cuts risk more quickly and reliably than either method alone.

Dipto Chakravarty, Black Duck’s Chief Product & Technology Officer, noted that AI is reshaping both the pace and economics of vulnerability discovery and exploit development. He added that combining Mythos’s capabilities with Black Duck’s existing deterministic testing, remediation tools, and governance controls turns vulnerability discovery into tangible, measurable risk reduction, while giving enterprise security teams the speed, transparency, and auditability they require.

The post Black Duck Joins Project Glasswing to Strengthen AI-Era Software Security appeared first on IT Security Guru.

Check Point Software Technologies has announced it is integrating OpenAI’s Daybreak frontier AI models across its security platform, extending a partnership aimed at helping defenders detect, validate, and remediate cyber risk faster.

The move builds on Check Point’s existing collaboration with OpenAI through the Daybreak Defense Network, first expanded three months ago, and follows the company’s recent decision to join more than 100 technology and security firms in backing OpenAI’s call for a collective, global surge in cyber defense.

In a blog post announcing the expansion, Check Point Chief Technology Officer Jonathan Zanger said the work does not stop with previous milestones, arguing that security needs to keep adapting as new threats and attacker capabilities emerge, alongside evolving technology stacks and growing enterprise use of AI.

Zanger said the aim is to put OpenAI’s frontier cyber reasoning to work across the security lifecycle, combining it with Check Point’s own security intelligence, context, and enforcement capabilities so customers can move from large volumes of raw security data to validated risk, actionable decisions, and faster protection.

Four Areas of Integration

According to Check Point, the Daybreak models are being rolled into four parts of its platform:

• Agentic Exposure Validation: Within Check Point’s Exposure Management product, the models are being piloted inside a multi-agent pipeline that separates genuinely exploitable risk from theoretical findings, combining AI reasoning with Check Point’s security context to validate attack paths and prioritise remediation.
• Agentic Security Management: As part of Check Point’s autonomous, intent-driven approach to network security management, the models will help investigate potential attack paths, understand vulnerabilities and risky exposures, and identify appropriate fixes, reducing manual policy management.
• Autonomous Workspace Platform: Within Harmony, Check Point’s investigation pipeline correlates email, endpoint, mobile, and browser telemetry; the models are being applied to investigate malware behaviour, attacker techniques, and credential-abuse chains, aiming to deliver clearer verdicts and remediation guidance while easing the load on security teams.
• Vulnerability research: The models are being used to accelerate analysis of vulnerable code and patches, identify realistic exploitation paths and reach verified results faster, without relying on publicly available exploit code, which Check Point says should translate into faster protection against newly disclosed vulnerabilities.

Check Point said it is taking a phased approach to the rollout: some capabilities are already in production, others are in development and being tested with design partners, with more to follow as the underlying technology matures. The company said every deployment follows the same discipline: governing what the model can see, constraining what it can act on, and testing and verifying its output before allowing it to take on more work within approved security workflows.

A Two-Way Relationship

Zanger framed the OpenAI partnership as operating in two directions: Check Point uses frontier AI to strengthen how it defends customers, while also helping those customers adopt and use OpenAI’s technologies securely. He said both sides of that relationship are becoming more important as AI moves beyond answering questions towards writing code and operating autonomous enterprise agents.

“Our goal is to give organizations the confidence to embrace what AI makes possible while staying protected against evolving risks,” Zanger said, adding that the partnership is intended to put frontier AI to work for defenders while helping customers deploy it safely themselves.

The announcement is the latest sign of security vendors racing to embed frontier AI reasoning models directly into detection, validation and remediation workflows, as both defenders and attackers increasingly turn to AI to gain an edge.

The post Check Point Brings OpenAI’s Daybreak Models Into Its Security Platform to Speed Up Threat Validation and Remediation appeared first on IT Security Guru.

Artificial intelligence is changing vulnerability discovery. At OpenSSL, we are seeing that change first-hand. A year ago, our security address received around nine separate reports and enquiries a month. It now receives around 70. AI tools can examine source code and identify potential security issues at a scale that would previously have required significant human effort.

In many ways, that’s positive. Finding vulnerabilities is an essential part of making software more secure. But there is another side to this that deserves much more attention. Every vulnerability report has to go somewhere.

Someone needs to assess whether the issue is genuine. If it is, engineers need to understand its severity, develop a fix, test that fix and manage disclosure appropriately. AI can increase the speed at which potential problems are discovered. It does not automatically increase the number of experienced engineers available to deal with them. That imbalance could become a serious issue for open-source security.

Finding a vulnerability is only the beginning

There is an understandable tendency to treat vulnerability discovery as the success story. An AI system finds something humans missed. That makes a compelling headline. But identifying a potential weakness and resolving it are very different tasks.

A report might represent a serious vulnerability. It might be something already understood. It might be technically correct but have limited real-world security impact. It might simply be wrong. Working out which of those is true requires expertise. Then, if there is a genuine vulnerability, somebody has to fix it.

Over the past 12 months we received a little over 400 vulnerability reports. 43 resulted in a published CVE. Roughly one in ten. The other nine still had to be read, understood, reproduced where we could, and answered. A report that turns out not to be a vulnerability consumes much the same expert attention as one that is — sometimes more, because establishing that something cannot be exploited is often harder than confirming that it can.

For a commercial software company with large security teams, increasing the number of reports may be manageable. For an open-source project with limited resources, a sudden increase can create a very different problem. The technology for finding possible vulnerabilities is becoming cheaper and more accessible. However, the expertise required to investigate them is not.

Businesses depend on projects they may barely know exist

This connects to a much older problem with open-source. Most technology companies know they use open-source software. What is less clear is whether they understand exactly which projects their products and services depend upon. That distinction matters.

Open-source components can sit deep inside software stacks. They work quietly, so organisations may have little reason to think about the people maintaining them. Then something goes wrong.

Heartbleed was an important moment for OpenSSL because it exposed the gap between the importance of open-source infrastructure and the resources available to support it. The industry responded. Investment increased and organisations began paying much more attention to the sustainability of critical open-source projects.

My concern is that some of those lessons are beginning to fade, and AI could make the consequences of that particularly visible.

AI changes the economics of vulnerability discovery

There is an asymmetry developing. The cost of searching code for potential security weaknesses is falling. The volume of reports can therefore rise significantly. But the other side of the process remains stubbornly human. Experienced engineers still need to understand the code. They need to judge whether the finding matters and decide how it should be fixed without creating another problem somewhere else.

Those people are a scarce resource. This means the question organisations should be asking about AI and cybersecurity isn’t only: “What can AI find?” It should also be: “Who is going to deal with everything it finds?”

For open-source projects, that leads directly to questions about sustainable funding. If businesses depend on a project as part of their critical infrastructure, supporting the health of that project should be viewed as part of resilience, not philanthropy.

Regulation only gets us part of the way

Governments are understandably looking at how regulation can improve cyber resilience. That matters, but regulation cannot maintain software. Europe provides some interesting examples of a different approach. OpenSSL Foundation has received support from Germany’s Sovereign Tech Agency, which invests directly in open digital infrastructure.

That recognises something important: if technology is critical to the functioning of the digital economy, somebody needs to invest in the people maintaining it. I’d like to see more of that conversation in the UK. Cyber resilience isn’t only about telling organisations what standards they should meet. We also need to consider the health of the technology underneath the services we’re trying to protect.

Organisations need to know what they depend on

There is something businesses can do immediately. Understand your open-source dependencies. If a critical vulnerability appeared tomorrow in a project your organisation relies on, could you identify where that software was being used?

Would you know which products and services were affected? Would you know who maintains the project? And would you have any relationship with the community responsible for fixing it? If the answer is no that is a resilience gap.

Organisations don’t necessarily need to contribute code themselves. There are other ways to support projects, including funding, engineering resources and participation in the communities maintaining the technology they depend upon. The important shift is recognising open source as infrastructure rather than free software that simply appears.

We need to talk about the people behind the code

AI will continue getting better at analysing software. That’s exciting, and it has the potential to make technology significantly more secure. But more findings do not automatically produce more security. The benefit comes when we have the expertise and resources to act on what those tools discover. That makes this a human question as much as a technology question.

How do we sustain the communities maintaining critical open-source infrastructure? How should businesses support the projects they depend on? What happens when vulnerability discovery accelerates faster than our ability to respond?

The post AI is finding vulnerabilities faster. Who is funding the people expected to fix them? appeared first on IT Security Guru.

An AI-assisted platform has been used to test whether Viasat’s satellite communications links can meet operational thresholds under interference and adversarial jamming. 

Announced this month, the work with Atalanta has renewed scrutiny of the vulnerabilities exposed by Russia’s 2022 attack on Viasat’s KA-SAT network, which disrupted communications across Ukraine and several European countries. 

Gil Baram, PhD, is a cybersecurity strategy and policy researcher specialising in cyber warfare, intelligence and space security. She is a Senior Lecturer and Associate Professor at Bar-Ilan University and formerly led the Cyber and Space Research Group at Tel Aviv University. 

In this exclusive interview for the IT Security Guru conducted by the Cyber Security Speakers Agency, Gil explains how states use cyber operations to compete below the threshold of open war, why long-life satellites present a distinct security challenge, and how an attack on space-based communications could disrupt everyday life on Earth. 

How have state cyber capabilities blurred the threshold between strategic competition and armed conflict? 

Gil Baram: “I think the greatest change that cyber warfare and cyber capabilities have created is that states can compete without crossing this imaginary red line: if you cross it, then it’s a war. 

“They don’t do that, but they still compete on and on, causing damage to one another, disrupting civilians’ lives and interfering in elections. But still, that doesn’t lead to an open war. That’s something very unique to this type of technology.” 

Why does securing space-based infrastructure present a different cyber challenge from defending terrestrial systems? 

Gil Baram: “When we talk about that, the first thing we have to have in mind is the distance. For many years, security contractors were the ones building big satellites for states and launching them. Have it in mind that these satellites should be in space for 20 or 25 years, and that’s a lot. 

“If you have a cyber risk, it’s very hard to patch a system that is flying far away from you and that was launched a decade ago. It created a lot of questions: how do you deal with cyber security in space? 

“In the past, we started seeing what we call the new space: startups and technology companies that started building satellites and building space capabilities. These are not the big security contractors. 

“We started seeing that these companies do have security by design, or cyber security by design, when they design their products and before they’re launching them to space. That’s a big shift we’ve seen today.” 

How exposed is civilian life to a successful cyberattack on satellite communications infrastructure? 

Gil Baram: “My feeling is that sometimes we don’t realise how much we depend on space capabilities, even for chatting today, using our internet or our cell phones, and navigating with GPS. Many things that we take for granted couldn’t happen without satellites and space satellite communication. 

“I’ll give just one example. In the first day of the Russian invasion to Ukraine, the Russian conducted a cyberattack against Ukraine satellite communication capabilities. 

“The reports coming out from Ukraine: they didn’t have the needed communication at the very critical moments or hours of the beginning of the war. You can damage satellite communication and then hurt an entire country. 

“The main thing I hope audiences will take, and that’s my passion and goal, is to understand or realise that cyber, AI and cyber threats are relevant to our everyday lives. It’s not something that is out there, and not just something that we read in the news, but it’s relevant for everyday lives. 

“We don’t have to be technical people in order to understand that and in order to protect ourselves.”

The post Q&A: Viasat Tests Satellite Resilience With AI as Cyber Expert Warns an Attack Could ‘Hurt an Entire Country’ appeared first on IT Security Guru.

KnowBe4 is set to explore the growing challenge of determining what organisations can trust in the age of AI at an exclusive cybersecurity briefing during Leeds Digital Festival 2026.

Taking place on 1 October, CyberSecure Leeds: Who Do You Trust Now? Human Judgement in the Age of AI will examine how the growing role of AI agents in business processes is changing the security landscape and creating new questions around identity, decision making and human oversight.

At the centre of the event will be an unusual real-world experiment from Javvad Malik, lead CISO advisor at KnowBe4, which started with a disputed parking ticket. During the experiment, Malik explored how AI could be used to help create convincing documentation and establish a digital presence that automated systems subsequently treated as legitimate, all within 48 hours.

While the experiment started with a relatively everyday scenario, KnowBe4 believes it highlights a much broader security concern. As organisations integrate AI into workflows, approvals and other business processes, systems may be placed in a position where they are effectively helping to determine what is legitimate and what should be trusted.

This becomes particularly significant as the digital workforce expands beyond human employees to include AI agents capable of accessing information, interacting with applications and influencing decisions.

“AI is rapidly becoming part of how organisations decide what is real, what is legitimate and what should be trusted,” said Malik.

“The problem is that AI can be extremely convincing without necessarily being right. As organisations give these systems greater influence over business processes, we need to think carefully about where human judgement fits and how attackers could exploit that trust. The workforce is now a combination of people and AI agents, and our approach to security has to reflect that reality.”

Phishing Beyond the Inbox

The briefing will also look at how the social engineering landscape is changing as attackers adopt AI-generated content and increasingly target users across multiple communications channels. James Dyer, head of threat intelligence at KnowBe4, will provide an update on current phishing activity, including the growth of phishing-as-a-service and the ways attackers are adapting their techniques to bypass existing security controls.

The session will examine why organisations can no longer treat phishing solely as an email security problem, with social engineering attacks capable of moving across different platforms and exploiting the trust employees place in digital communications.

The event coincides with the beginning of Cybersecurity Awareness Month and forms part of Leeds Digital Festival, which runs from 21 September to 2 October with more than 200 technology events taking place across the region.

The KnowBe4 briefing will run from 8:30am to 11:00am at the company’s Leeds office and is aimed at CISOs, IT leaders, security practitioners and business decision makers.

Attendance is free but limited to 30 places, with advance registration required through the Leeds Digital Festival event programme. Interested delegates can register here: https://web.cvent.com/event/96b273cf-5129-4734-b54c-0dfd0a38aa61/summary

The post KnowBe4 to Put AI Trust to the Test at Leeds Digital Festival appeared first on IT Security Guru.

The perception that compliance is a once-a-year scramble is out of date, according to a new survey of 201 security and compliance practitioners published by Pentest-Tools.com. The research finds that continuous compliance has effectively already arrived inside most organisations, but the automation needed to support it has not.

The study, carried out in July 2026 and drawn from IT managers, compliance and GRC leads, security engineers, DevSecOps professionals and security specialists, set out to test how organisations maintain certifications such as ISO 27001 and SOC 2. Rather than surveying executives or auditors, Pentest-Tools.com went directly to the practitioners responsible for the day-to-day evidence work.

Assessment cycles have quietly gone continuous

According to the findings, 60.2% of respondents now formally assess the effectiveness of their security controls at least monthly, with 37.8% doing so continuously and a further 22.4% monthly. Just 9.5% still work to an annual assessment cycle, and no organisation in the sample assessed controls less often than every six months.

That cadence has outpaced the rate at which the underlying environments actually change. More than half of respondents (56.8%) said their production systems change monthly or less often, meaning many organisations are now checking their controls as frequently as, or more frequently than, the systems themselves are updated.

The report suggests two possible explanations that likely both hold true: some organisations have deliberately decoupled control validation from the deployment cycle, treating assessment as an always-on discipline, while others are responding to external pressure, such as the rising volume of disclosed and actively exploited vulnerabilities, which can leave an unchanged system newly exposed.

The bottleneck is evidence, not policy

A central finding of the report is that audit delays are rarely caused by outstanding policy or documentation work. Instead, when asked to name their biggest audit preparation bottlenecks, respondents pointed overwhelmingly to operational and technical constraints: getting time and input from technical teams (50.7%), obtaining the required technical evidence in time (42.8%), coordinating responses across multiple teams (36.3%), and resolving outstanding security findings before the audit (34.3%). Only 3.5% of respondents reported no significant bottlenecks at all.

The same pattern emerged when practitioners were asked which parts of maintaining compliance evidence consume the most time. Detection, defined as collecting evidence from available sources, topped the list at 45.8%, ahead of validating that findings are real and repeatable (38.8%) and demonstrating that fixes have held (36.8%). Notably, only 2% of respondents said maintaining compliance evidence was not a significant workload.

“Security teams mostly already have the information auditors need,” the report states. “The bottleneck is turning what they know into what they can show.”

Automation hasn’t kept pace with ambition

While 80.6% of respondents said they collect compliance evidence throughout the year, only 38.3% said they rely primarily on automated tooling. A larger group, 42.3%, said they maintain continuous evidence collection but depend on significant manual consolidation to keep it usable, while 14.9% still assemble evidence largely before audits.

The duplication compounds further once multiple frameworks are involved, which the survey found to be the norm: almost nine in ten respondents maintain more than one compliance framework. Just 6.5% said tooling or templates handle most of the evidence mapping between frameworks. The majority, 63.2%, said they partially remap evidence by hand, and a further 25.9% said most evidence is manually re-documented separately for each framework, meaning the same underlying proof is often reproduced multiple times for different audits.

Certification is trusted, but seen as time-limited

The survey also examined how much confidence practitioners place in certification itself. While 93% of respondents said certification reflects their organisation’s security posture to some degree, only 51.2% believe it does so continuously. A further 41.8% said it is accurate only immediately after an assessment, implying that confidence in the badge fades over the certification cycle for a substantial share of practitioners.

That belief was closely linked to confidence in a hypothetical surprise audit. Among practitioners who see certification as continuously accurate, 59.2% said they would be very confident demonstrating control effectiveness in a next-day audit. Among those who believe it is only accurate immediately after assessment, that figure fell to 16.7%.

Most organisations catch failures before auditors do

One of the more positive findings challenges the assumption that organisations typically discover control failures during audit preparation. Only around 12% of respondents said they first learn of failures during audit preparation or the audit itself. The majority instead surface issues through continuous automated monitoring or scanning (44.3%) or routine internal security reviews (24.4%).

The report notes a maturity gradient by organisation size: among organisations with fewer than 100 employees, 31.3% rely on continuous automated monitoring to surface failures, compared with roughly half of organisations with more than 1,000 employees. Pentest-Tools.com suggests this gap reflects tooling maturity rather than headcount, and argues that automated discovery is one of the few maturity gains smaller teams can achieve without adding staff.

Practitioners want less manual work, not more integrations

When asked to rank which factors matter most when evaluating tools or workflows for audit readiness, continuous, automated evidence generation was the clear priority, ranked most important by 24.4% of respondents and least important by just 3%. By contrast, integration with existing GRC and ticketing tools was ranked most important by only 12.9% and drew the second-highest “matters least” score of any factor.

The report highlights an apparent contradiction: in open-ended responses, automation and integration together accounted for 33.5% of all suggestions when practitioners were asked what they would change about their compliance workflows. Pentest-Tools.com concludes that respondents want a single, continuously current evidence base rather than additional point-to-point connections between tools that still require manual consolidation.

What this means for security and compliance teams

Taken together, the findings paint a picture of a discipline whose ambitions have shifted toward continuous validation faster than its supporting tooling has evolved. Assessment cadences have tightened, evidence volumes have grown, and the number of frameworks organisations must satisfy continues to expand, but much of the resulting workload still falls to technical teams manually collecting, validating and re-mapping proof by hand.

Adrian Furtuna, founder and CEO of Pentest-Tools.com, said the survey was intended to capture the experience of the people actually doing the work rather than executives or auditors. “The bottleneck in modern compliance isn’t policy or paperwork,” the report states. “It’s producing technical evidence that security controls actually work, at the pace environments change, without burning out the engineering teams who hold that evidence.”

The full report, “The audit bottleneck isn’t policy. It’s proof.,” is available from Pentest-Tools.com.

The post Compliance teams have gone continuous, but their evidence-gathering hasn’t caught up appeared first on IT Security Guru.

By Aimee Steele, threat intelligence analyst at Talion Cyber Security

Last month, the UK’s National Cyber Security Centre (NCSC) issued an advisory around a new phishing campaign targeting organisations in the West that was being carried out by the Russian state-sponsored threat actor known as Laundry Bear.

The campaign, saw the threat actors exploiting a zero-day vulnerability in Zimbra Collaboration Suite in order to compromise networks, before gaining persistence, accessing emails, stealing sensitive data and conducting espionage against organisations using vulnerable Zimbra Mailservers.

The advisory from the NCSC was issued in conjunction with advisories from 15 other countries, including the US, Finland, Australia, Denmark and France, and it advised that the threat actors were targeting critical industries in order to steal sensitive information with Russian government backing.

Phishing is nothing new and is regularly cited as the most prevalent method used within cyberattacks. However, what made this attack stand out and warrant such a coordinated international response, was its deviation from conventional phishing techniques, amounting to a markedly sophisticated zero-click campaign.

Successful exploitation did not rely on users clicking a link or opening an attachment; victims simply needed to open an email.

Typical phishing attacks rely on user interaction. The victim might click on a malicious link or open an attachment. The threat actor might leverage phishing techniques to deploy malware, infiltrate systems or networks and establish backdoors for persistent access. Even at the most basic level, attackers might simply use stolen credentials to access private user accounts, such as a bank account, to commit financial fraud.

In this instance, successful exploitation did not depend on outright deception; there were no features that might typically raise a user’s suspicion. The victim didn’t need to do anything at all.

So, how did the attackers get in?

The campaign

The threat group was able to exploit a zero-day XSS vulnerability in the Zimbra Collaboration Suite (CVE-2025-66376), the exploit having been embedded in the HTML body of the message and triggered upon opening or previewing. Emails were issued from either Proton Mail accounts or previously compromised email addresses.

Successful exploitation gave attackers access to the webmail server, allowing them to establish persistence, access users’ mailboxes and steal sensitive emails and authentication data.

From there, the stolen information could be used to conduct cyber espionage, gather intelligence on government and critical infrastructure organisations, and potentially facilitate further compromise of connected systems. Attackers were also able to leverage their access to target specific victims in spearphishing attempts. In instances such as this, having access to contextual information can add an additional layer of legitimacy to communications.

Protecting against Zero-Click attacks

Following the discovery of the vulnerability, Zimbra released security updates to address the flaw, and any organisation using affected versions of Zimbra Collaboration Suite should ensure these patches have been applied as a priority. If this is not immediately actionable, it is advisable for organisations to implement another suitable mail client in the meantime.

However, patching alone should not be considered sufficient. Organisations should also review logs, monitor for indicators of compromise, investigate unusual authentication activity and assess whether attackers may have gained access before the vulnerability was remediated. Multi-factor authentication, network segmentation and continuous monitoring can also help limit the impact if a compromise does occur.

Attacks like these also reinforce the importance of detection and response, which can help spot malicious access quickly, before attackers have an opportunity to compromise networks and access data.

Zero-click exploits present a greater challenge to the individual user than typical phishing attempts. However, they can still play an important role in limiting the impact of such attacks by remaining vigilant for unusual login alerts, unexpected password reset notifications or signs that an account has been accessed without permission.

Users should also ensure multi-factor authentication is enabled wherever possible, use strong and unique passwords, and be alert for any follow-on phishing attempts, as attackers will often use information stolen during an initial compromise to target victims further.

Lessons for organisations

Zero-click attacks are particularly deceptive because they present without the typical warning signs users are trained to look out for in social engineering and phishing scams.

This recent zero-click campaign by Laundry Bear begs the question as to whether current user awareness training is particularly affective in the prevention of successful phishing attacks, which are becoming increasingly sophisticated.

In general, rapid patching supports robust remediation strategy and is fundamental if organisations wish to prevent exploitation, rather than react to it. In addition, continuous monitoring and effective detection and response capabilities are fundamental when reacting efficiently to zero-clicks.

Considering the severity of the attack, any organisation using Zimbra Collaboration Suite should ensure they have applied the latest security updates as a priority.

Organisations should also assume patching alone is insufficient, and review logs, authentication activity and other indicators of compromise to determine whether they were breached before the vulnerability was remediated.

Overall, attacks like these require a layered approach to security, helping identify attackers and potential threats before organisations are compromised further.

The post Protecting Against Zero-Click Attacks appeared first on IT Security Guru.

Cybersecurity firm Huntress has uncovered a wave of malicious installations of ScreenConnect, a widely used remote-support tool, that spread between machines without any further action from a victim or an attacker, a self-propagating attack chain researchers likened to a computer worm.

In a blog post published this week, Huntress said its Security Operations Center (SOC) had flagged the same unusual pattern of activity across several unrelated customer environments in late August. Investigators later found the incidents were connected by a shared attack chain built around modified, or “rogue,” copies of ScreenConnect, a legitimate remote-access product made by ConnectWise that IT teams and help desks use to support end users remotely.

A familiar scam, an unfamiliar twist

Each incident Huntress examined began with social engineering. In one case, a victim ran Microsoft’s built-in Quick Assist tool after being convinced, likely through a fake tech-support call, that their computer had been compromised, a well-worn tactic in tech-support scams. In another, a user searching online for a Geek Squad refund form was instead led to download and run a bogus ScreenConnect installer.

Once the rogue ScreenConnect client was installed, each infected machine began repeatedly launching the Windows Script Host process to run a sequence of four VBScript files, named simply 1.vbs through 4.vbs. Huntress said this behaviour, along with a matching persistence mechanism disguised as a “WindowsServiceHost” registry entry, appeared consistently across every incident, despite the organisations involved having no obvious connection to one another.

According to Huntress’s analysis, the four scripts work in stages. The first profiles the infected machine, checking whether ScreenConnect is already installed, cataloguing which security products are running, and confirming the system has enough memory to plausibly be a real computer rather than a malware analyst’s virtual machine. Based on that profile, later scripts pull down and decrypt additional payloads, which can include a backdoored ScreenConnect client, tools for privilege escalation and persistence, or a bundle containing tunnelling software and a cryptocurrency miner.

Turning victims into distribution points

The most striking element of the campaign, Huntress said, is how it spreads. Buried inside the backdoored ScreenConnect client is code that watches for new incoming remote-support sessions. When a new connection appears, the infected client automatically packages up the same four VBScript files and pushes them to the newly connected system, triggering the same infection chain there too.

In practice, that means a legitimate support session, a technician or help desk agent remotely connecting to a compromised machine to assist a user, could result in the malware spreading onward to the technician’s own environment, with no additional phishing or social engineering required at that stage. Huntress said the infected client keeps track of which sessions it has already targeted, but drops that record once a session ends, allowing the same host to be reinfected on a later reconnection.

Huntress also observed secondary remote-access tools, including UltraViewer, deployed on some compromised machines, suggesting the attackers were establishing multiple footholds in case one was discovered and removed.

Researchers point to LLM-assisted development

While unpacking the scripts, Huntress researchers noted a comment embedded in one of the VBScript files that appeared to explain, in unusually plain language, how to parse an encryption key out of a configuration file, the kind of explanatory comment researchers said was consistent with code generated with the help of a large language model.

What organisations should do

Given the depth of access the malware can obtain, including attempts to disable Microsoft Defender reporting and bypass User Account Control, Huntress recommended wiping and reimaging any confirmed infected machine from known-clean media rather than cleaning it in place.

The firm urged administrators to scrutinise any on-premises ScreenConnect deployments and to check ScreenConnect server audit logs for RunFiles or RanFiles entries showing scripts executed from a “Guest” process, which it said should be treated as an immediate red flag. Huntress cautioned that the specific filenames associated with the campaign may change over time, and that any unexpected Windows Script Host or PowerShell activity tied to ScreenConnect sessions should be investigated.

Huntress said it is in direct contact with ConnectWise, ScreenConnect’s maker, and continues to monitor the activity. The report includes a full set of indicators of compromise, including file hashes and command-and-control infrastructure, for defenders to check against their own environments.

Remote monitoring and management tools like ScreenConnect have been among the most abused categories of software this year, according to Huntress, which has previously documented social-engineering campaigns using the same class of tool to gain initial access to victim networks. What sets this campaign apart, researchers said, is the addition of automated, worm-like propagation on top of an already common attack vector.

The post Hijacked ScreenConnect Installs Are Spreading Malware Like a Worm, Huntress Warns appeared first on IT Security Guru.

KnowBe4 has appointed cybersecurity channel veteran Kurt Mills as its new channel chief, as the company looks to strengthen its global partner ecosystem and expand its channel routes to market.

In the role, Mills will lead the evolution of KnowBe4’s global partner programmes and operations, with responsibility for supporting the company’s relationships across VARs, MSPs, systems integrators, distributors and technology partners.

Mills brings more than 25 years of channel experience to KnowBe4, having previously held executive channel and partner sales positions at Check Point Software Technologies, Mimecast, FireMon and Blue Coat Systems.

Throughout his career, he has focused on aligning partner strategies with wider business and revenue goals, building partner networks and channel teams through periods including IPOs, acquisitions and rapid market expansion.

“Partner ecosystem growth is central to our mission, and Kurt’s appointment reflects our continued investment in that strategy,” said Bryan Palma, CEO of KnowBe4. “Kurt’s deep channel expertise, proven ability to scale high-performing teams, and track record of building long-term partner value will be instrumental as we expand our market reach.”

Mills’ appointment follows a number of additions to KnowBe4’s channel leadership team, including Neill Burton and John Noha, who have both joined as vice president of channel to support the company’s global initiatives.

The expanded leadership team will focus on providing partners with the programmes, tools and resources needed to identify new opportunities and support customers throughout the security lifecycle.

KnowBe4 said the investment reflects the increasingly important role partners play in its global growth strategy and in delivering its digital workforce security offering to organisations worldwide.

“Joining the KnowBe4 team at this critical point in the company’s growth is an incredible opportunity,” said Kurt Mills, newly appointed channel chief at KnowBe4. “I’ve built my career on creating focused, measurable, and engaged teams that drive mutual success for organizations and their partners. I am thrilled to team up with everyone at KnowBe4 to scale our partner programs, deepen our relationships across VARs, MSPs, SIs, distributors, and tech partners, and fuel our next phase of global growth.”

The post KnowBe4 Names Kurt Mills as Channel Chief to Lead Next Phase of Partner-Led Growth appeared first on IT Security Guru.

A newly identified phishing-as-a-service kit is being used to hijack Microsoft 365 accounts by stealing victims’ active login sessions rather than their passwords, according to new research from cybersecurity firm Huntress, a technique that allows attackers to walk straight past multi-factor authentication (MFA) without ever needing to guess, crack, or bypass it.

The kit, dubbed “Knight Office” by researchers, came to light after Huntress’s Security Operations Centre investigated suspicious sign-in activity on a customer’s Microsoft 365 account in August. While tracing the source of the intrusion, analysts found the attacker’s own operator console, a slickly built dashboard, complete with a Cloudflare Turnstile bot-check and real-time visitor statistics, used to manage victims and harvested logins from a single screen.

Huntress was careful to distinguish this operator-facing panel, which gives the attacker a live view of victims and their stolen data, from the underlying phishing kit code itself, which handles the victim-facing side of the attack, such as the fake login pages used to trick targets.

How the attack works

According to Huntress, the campaign begins with a fake email styled to look like a DocuSign signature request, complete with an urgent subject line pressuring the recipient to act. In a twist designed to dodge suspicion, the emails are “self-spoofed” and forged so that they appear to come from the recipient’s own email address.

Clicking the link in the email sends victims through a chain of redirects, including via the legitimate Monday.com work-management platform and a compromised Joomla website, designed to obscure the final destination from email security scanners. Victims land on a convincing fake Microsoft page that presents them with what looks like a normal “device login” code, mimicking the legitimate process Microsoft uses to sign into apps on other devices.

Once a victim enters the code and completes the Microsoft login, including approving the MFA prompt on their phone, exactly as they would for a genuine sign-in — the attacker’s infrastructure silently intercepts and captures their live session. That stolen session is instantly usable, letting the attacker access the account as though they were the legitimate user, with no password and no MFA prompt of their own required.

Huntress said its telemetry showed the stolen tokens being reused from data-centre hosting infrastructure, and that because no password was ever entered incorrectly, standard password-based detection alerts never fired.

Attackers dug in for the long haul

Huntress’s investigation found that the attacker didn’t stop at the initial break-in. After gaining access, they registered a rogue, attacker-controlled device against the victim’s Microsoft Entra ID (formerly Azure AD) tenant and bound a Windows Hello for Business (WHfB) passwordless credential to the compromised account, effectively planting a backdoor that would let them log back in even after the original stolen session was revoked.

The technique, researchers noted, turns a feature designed to make sign-in more secure for legitimate users into a persistence mechanism for attackers.

Scale of the campaign

Huntress said the same operator console has been linked to at least nine confirmed phishing attacks on Microsoft 365 and Google Workspace accounts within its own customer base over a two-week period. Separately, the firm said more than 700 emails matching the same lure and template have been reported by users through its security awareness training platform since April, suggesting the campaign has been running for months at considerably larger scale.

Variants of the lure’s subject line spotted by researchers include messages disguised as voicemail notifications and shared-document alerts, several of which substitute the lowercase letter “l” for “i” in words such as “Important” and “Signature” — likely an attempt to dodge spam filters that scan for exact keyword matches.

Part of a wider shift away from passwords

Knight Office is the latest in a string of phishing kits Huntress has tracked that focus on stealing session tokens or OAuth access tokens rather than credentials, following similar findings around kits called EvilTokens and Kali365. Security researchers have warned that this style of attack, known as adversary-in-the-middle (AiTM) phishing, is becoming increasingly popular precisely because it renders traditional password hygiene and even MFA far less effective as standalone defences.

Huntress recommended that organisations look beyond failed login attempts and password-spray alerts when hunting for this kind of activity, and instead watch for unexpected post-MFA authentication events from unfamiliar devices or callback proxies, review newly registered Entra ID devices and WHfB credentials, and promptly revoke unauthorised authentication methods when found.

Indicators of Compromise

Huntress has published a full list of indicators of compromise associated with the campaign, including the IP address of the phishing control panel (104.37.188[.]94), the domain hosting it (idoej[.]com), and more than two dozen lookalike phishing domains using the .vu top-level domain.

The post New ‘Knight Office’ Phishing Kit Steals Microsoft 365 Logins Without Touching a Password appeared first on IT Security Guru.

Application security vendor Black Duck has launched its Signal vulnerability scanning engine as an MCP server in the Claude Directory, giving developers using Anthropic’s Claude Desktop a way to check code for security flaws without switching tools.

The integration is built on the Model Context Protocol (MCP), the open standard that lets AI assistants like Claude call out to external services and pull structured data back into a conversation. Through it, Black Duck’s Signal Code Analysis engine can scan git diffs, individual files, or whole codebases for vulnerabilities directly from within the Claude environment developers are already using to write and refactor code.

Under the hood, source code submitted for a scan is sent to Black Duck’s cloud-based analysis service for processing. The results then come back as MCP resources, structured data that Claude can read and reason over, allowing it to explain identified risks and suggest remediation steps in plain language rather than simply returning a raw findings report.

The launch reflects a wider shift in how security vendors are approaching AI-assisted coding. As tools like Claude speed up how quickly developers can write and ship software, security teams are under pressure to embed checks earlier in the process rather than relying on scans that happen after code has already been merged. Black Duck is positioning Signal as a way to close that gap by putting vulnerability detection at the point of code generation itself.

Dipto Chakravarty, Chief Product & Technology Officer at Black Duck, framed the move as a response to the pace at which AI coding tools now operate. “security keeps pace with how fast teams are building,” he said of the aim behind bringing Signal into the Claude Directory.

Signal is available now through the Claude Directory listing, with Black Duck directing prospective users to speak to a company representative to get set up. The release follows Black Duck’s broader push into AI-focused application security tooling, part of a growing trend among established AppSec vendors to adapt existing scanning capabilities for workflows increasingly driven by AI coding assistants rather than traditional IDEs.

It also underscores the growing role of MCP as connective tissue between AI assistants and specialist enterprise tools. Since Anthropic opened up the protocol, a steady stream of security, development and productivity vendors have released their own MCP servers, letting Claude act as a front end for capabilities that would otherwise require developers to leave their AI workflow entirely.

The post Black Duck brings AI-powered vulnerability scanning into Claude with new Signal integration appeared first on IT Security Guru.

Security researchers at Huntress have discovered a malware campaign that tricks victims into installing a real, fully functional copy of Exodus, a popular cryptocurrency wallet application, only to disable it so it can never actually be opened, using it instead as cover for a hidden spying tool.

The firm said it identified four separate organisations compromised between late July and mid-August 2026, three of them within an 85-minute window on a single day, using a version of the malware built the day before it was deployed.

According to Huntress, victims were lured into opening what appeared to be a work document or a routine software update. In practice, the files silently downloaded a Windows installer that presented itself, falsely, as an Apple “Background Service.” The installer placed a genuine, largely unmodified copy of the Exodus wallet, version 24.33.4, onto the victim’s machine. Of nearly 2,000 files bundled with the wallet, researchers found that only three had been altered.

One of those changes prevented the wallet from ever displaying a window, meaning it never appeared in the taskbar and could not be closed or interacted with by the user. The other two altered files formed a loader that decrypted and ran a separate 10-megabyte payload directly in the computer’s memory, without ever writing it to disk.

That hidden payload, researchers said, is a modular remote access trojan (RAT) with six separate capabilities: remote command execution, file browsing and transfer, a hidden virtual network connection (VNC) allowing an attacker to view and interact with the victim’s desktop unseen, a SOCKS proxy that could turn the infected computer into a relay for further attacks, a scripting engine, and a module built to steal saved passwords, cookies, and browser data from Chrome, Edge, and Firefox.

Despite the crypto wallet disguise, Huntress said the malware does not target wallet data, seed phrases, or cryptocurrency funds directly. Researchers instead described the campaign as being built for broad, hands-on remote access and credential theft, with the wallet acting purely as camouflage.

Because most of the installed software is genuine, unaltered Exodus code communicating with real Exodus servers, the tampered installer registered zero detections across 76 antivirus engines on VirusTotal at the time of testing, according to Huntress.

The malware also avoided using attacker-owned servers to communicate. Instead, researchers found it checking in with Microsoft’s Azure Table Storage service, a technique that allowed its network traffic to blend in with ordinary, legitimate cloud activity rather than triggering alerts tied to suspicious domains.

To maintain access, the malware created a scheduled task that silently relaunched the invisible wallet every hour. Huntress also identified a second, earlier-observed task that repeatedly reset a victim’s corporate proxy settings, a mechanism apparently built specifically to keep the malware’s outbound traffic from being blocked or inspected by network security controls.

Huntress said the operation shows signs of a single toolchain behind both the delivery mechanism and the trojanised wallet, including a shared code-obfuscation technique and an npm software package impersonating Intel. Researchers recommended that any organisation identifying these artefacts treat the affected host as a full interactive compromise, revoke active browser sessions, and rotate credentials rather than only resetting passwords.

Huntress’s full technical writeup, including indicators of compromise, is available at: The Crypto Wallet That Never Opened: Tampered Exodus Installer Hides a Modular RAT

The post Fake Software Update Installs a Real Crypto Wallet – Rigged So It Can Never Open appeared first on IT Security Guru.

Cybersecurity researchers at Huntress have uncovered a phishing campaign that abuses Faronics Deploy, a legitimate endpoint management platform used by businesses, schools, and government offices to remotely install software and run scripts across their networks.

According to the security firm, threat actors sent victims phishing emails disguised as invoices, tax documents, financial records, and event invitations. Clicking through led to convincing fake pages, including bogus “document downloading” screens and spoofed Adobe plugin update prompts, that ultimately delivered a genuine, digitally signed Faronics Deploy installer disguised as the expected file.

Once installed, the software enrolled the victim’s device into a Faronics deployment controlled by the attacker, granting them the ability to remotely execute scripts with administrative privileges. Huntress said it repeatedly observed attackers using this access to run PowerShell commands that retrieved additional malicious scripts from external sources, including GitHub, which were then used to install ScreenConnect, a separate remote access tool, on the compromised machine.

Because the initial software was legitimate and properly signed, the activity was able to blend into normal business IT workflows and evade detection by traditional security tools, the researchers said.

Huntress reported observing more than 457 endpoints encountering Faronics-related phishing lures over a one-month period between July 21 and August 20. The company also identified visitor-fingerprinting techniques used to filter out security researchers and mobile users, serving them decoy content, including a fake website impersonating Eastern Bank, instead of the malicious download flow.

Huntress reported the activity to Faronics on August 5. The vendor has since implemented new measures to detect and block fraudulent account registrations, including those using impersonated organisations and compromised accounts, and has proactively contacted customers whose accounts may have been affected. Huntress said it observed a sharp decline in malicious activity involving the platform beginning around August 21.

Researchers noted that key forensic artefacts, including a log file named ScriptRunner.log and a deployment identifier known as the “ck” parameter, can help defenders trace and cluster malicious Faronics activity, even after the original attack scripts are no longer available.

Organisations that discover an unauthorised Faronics installation are advised to isolate the affected endpoint, review the ScriptRunner.log for remotely executed scripts and external URLs, check for secondary remote access tools such as ScreenConnect, and rotate credentials where interactive attacker access is suspected. Huntress said organisations that identify suspected abuse of the platform can report it directly to Faronics support.

Huntress’s full technical writeup, including indicators of compromise, is available at: Daisy-Chaining Trust: Investigating Faronics Deploy Abuse

The post Hackers Abuse Legitimate IT Management Tool to Sneak Into Business Networks appeared first on IT Security Guru.