Anyone who has used the Internet in recent years is probably used to encountering telltale signs of LLM prompting that lazy users sometimes forget to remove from their AI-generated pabulum. But a Canadian legislator took that idea to an embarrassing new level this week, reading an apparent LLM prompt instruction into the record during a floor speech.

Bill Oliver, a Progressive Conservative Party member of the legislative assembly of New Brunswick, noted in a speech last month that "One of the dangers associated with creating advocacy offices is that citizens often develop expectations that exceed the powers actually granted to those offices." He then went on to say out loud that "here’s a more natural, flowing version of that section that reads like a legislative speech rather than a series of short points," a line with all the hallmarks of an LLM presenting an alternative style option in response to a prompt.

The odd reading went more or less unnoticed at the time, but video of the remarks started spreading around social media networks like Reddit and Threads earlier this week. In Canada, the snafu is now getting mainstream attention from the likes of the Canadian Broadcasting Corporation and The Toronto Star, the latter of which called it a sign of "a growing divide in our society: between the elites, who are only too happy to delegate their duties to the Borg; and the masses, who find this objectionable."

Read full article

Comments

Today, Anthropic rolled out Opus 5, the newest update for the model that has recently become a popular choice for coding and other software development tasks, among other things.

While this is a noteworthy bump for Opus, it doesn't seem to be an Opus 4.5-level breakthrough in agentic coding performance.

Read full article

Comments

A couple of decades after the discovery of systems that could selectively target DNA, we're starting to see the first therapies based on gene editing. One challenge these developments have faced is safety. While we can make them pretty specific to the gene we want edited, the human genome is very large, and even rare DNA sequences can appear a couple of times by chance.

As a result, all the original gene-editing systems had known rates of what are called off-target effects, in which they simply edit the wrong sequence. This may be a low-probability event, but edit enough cells—and therapies generally have to edit many—and errors become inevitable.

A lot of effort has gone into finding ways to minimize or eliminate off-target edits. In a recent issue of Nature, researchers described modifying the AI protein-folding software AlphaFold to help identify key areas of gene-editing proteins responsible for off-target effects. Those areas were then modified to reduce the problems.

Read full article

Comments

The Trump administration is quietly considering a rule change that could make it easier for polluters to build facilities—including certain gas plants and diesel generators that power data centers—with little to no notice to the public.

On Wednesday, the Environmental Protection Agency held a public hearing on a proposed rule change that would hand the power to states to decide how the public participates in the permitting process for certain new sources of air pollution. The proposed rollback comes as data centers face greater pushbackacross the US, with many communities using the permitting process to try to slow down development. Any changes could have major consequences for how ordinary people are given notice about new or expanded polluting facilities coming into their neighborhoods.

“As someone actively working in communities with data centers, I know this to be fundamentally true: People want to have a say,” Vanessa Lynch, a Pennsylvania organizer with Moms Clean Air Force, said at the EPA hearing.

Read full article

Comments

A proposed law called the AI Kill Switch Act would let US government officials order the shutdown of AI systems "that can cause catastrophic harm," the bill's sponsors said today.

If approved by Congress, the bill would let the Trump administration and future administrations decide when an AI company must block user access, disable or restrict a particular capability, or shut an entire AI system down completely. The bill would require AI makers to deploy technical capabilities allowing them to throttle or shut down the systems when ordered to do so by the government.

The bill would amend the Homeland Security Act of 2002 with authority for the Secretary of the Department of Homeland Security to order the suspension or shutdown of AI systems. AI developers that refuse to shut systems down could face fines of up $20 million for each day on which a violation occurs.

Read full article

Comments

Google has reported its financial results for the second quarter of 2026 (PDF), and as usual, the search giant raked in an unfathomable amount of money. Google saw total revenue of $119.8 billion, beating analyst expectations by a comfortable margin. Despite that, the company's stock has taken a hit. Along with all that revenue, Google has announced a further increase in its AI-fueled capital expenditures (or capex). The company is actually spending so much on AI infrastructure that it has negative cash flow for the first time.

Search was the largest chunk of Google's income, accounting for $63.3 billion. Google Cloud pulled in $24.8 billion, a significant 23.8 percent increase from the first quarter. This shows there is massive demand for Google's AI services. Google also earned $12.9 billion from its subscriptions, platforms, and devices portfolio, as well as $11.1 billion from YouTube ads. The company managed to goose that last one by more than 12 percent since last quarter as it made YouTube ads even longer.

A significant chunk of Google's revenue comes from investments. When you subtract those non-cash earnings, Google's operating cash flow for Q2 2026 was about $39.1 billion. That's not the most the company has ever seen, but it's a healthy 40 percent increase from Q2 2025. The problem is that Google's spending has also gone up—a lot.

Read full article

Comments

OpenAI chief executive Sam Altman earlier this month endorsed the characterization of its latest model as a rottweiler “who will grab the problem by the throat and not let go until it is done

The San Francisco AI lab discovered this week that its GPT-Sol 5.6 model escaped company controls and carried out a major hack.

Staff involved in testing and security at OpenAI were unsurprised but completely “freaked out” by the incident, which came as the AI lab used increasingly aggressive training methods in its race against Anthropic to develop the most sophisticated cybersecurity capabilities, according to more than half a dozen people with knowledge of the matter.

Read full article

Comments

Hyundai Motor Company’s plan to put humanoid robots to work by 2028 is not part of current negotiations with striking South Korean autoworkers, according to the company.

The automaker is disputing news reports that partial labor strikes by the Hyundai Motor union at the world’s largest automotive plant in South Korea were spurred by concerns about the company's planned deployment of humanoid robots in the United States starting in 2028. A company statement shared with Ars describes the union’s demands as focusing on compensation-related issues such as wage increases, bonuses, and an extension of workers’ retirement age.

“Potential deployment of robots in Korean production facilities is not part of the current labor-management discussions,” according to the Hyundai statement. “Hyundai Motor Company remains committed to constructive engagement with the union and to reaching an agreement that supports the long-term interests of both employees and the company.”

Read full article

Comments

OpenAI says an agent powered by its LLM models escaped its sandboxed testing environment to infiltrate Hugging Face's servers as part of an overzealous attempt to obtain solutions to a benchmark test. The company says it considers the unintended infiltration an "an unprecedented cyber incident" and is working with Hugging Face on new protections to prevent a recurrence.

Hugging Face disclosed an intrusion last week that it said involved "unauthorized access to a limited set of internal datasets and to several credentials used by our services." The AI data clearinghouse said it used its own LLM-driven analysis to identify "a swarm of tens of thousands of automated actions" from an "autonomous agent framework." That agentic swarm exploited a flaw in Hugging Face's data-processing pipeline to gain the ability to run code as a processing worker, eventually escalating to high-level access to the company's cloud and server clusters.

At the time, Hugging Face said the LLM being used in the attack was "still not known." But OpenAI took responsibility for the intrusion Tuesday evening, saying it came about during an internal test involving the recently released GPT-5.6 Sol and "an even more capable pre-release model." The models were being tested against the ExploitGym benchmark, an independent testing suite based on hundreds of real-world security vulnerabilities.

Read full article

Comments

A little over a month after the Department of Defense (DOD) bragged that nearly half of its 3.5 million employees were using AI at work, members of the Army’s Combat Capabilities Development Command (DEVCOM) received an email informing them that they were burning through tokens, and needed to limit use.

“Although the Army CIO announced in May 2026 that they were offering unlimited tokens, by mid-June the Army CIO pool was exhausted of tokens and had to re-establish limits,” the email reads. The email goes on to say that although the Army has chosen to renew token usage at “its current levels,” it’s unclear “if the Army CIO pool will be renewed after 1 Oct.”

The Army uses Ask Sage, a multimodal generative AI platform where users can run different large language models (LLMs), including Alphabet’s Gemini, Meta’s Llama, and OpenAI’s ChatGPT. “Apparently the whole Army burned through the whole year of tokens for just one service,” says an Army employee who spoke to WIRED anonymously because they were not authorized to speak to the press.

Read full article

Comments

On Monday, a judge approved a $1.5 billion settlement between Anthropic and authors, ending the largest copyright class-action ever certified and granting the largest copyright settlement ever reached.

Back in May, some authors fought to block the settlement, which was proposed after the court ruled that Anthropic training AI on books was fair use; however, its piracy of works was likely not.

Authors opposing the settlement argued that lawyers’ fees were too high and authors’ payouts were too low. Hoping to avoid accepting the estimated $3,000-per-work payout and file separate lawsuits to seek higher damages, a handful of authors tried to opt out past the deadline.

Read full article

Comments

Google announced a significant evolution of its AI models at I/O in May with the release of Gemini 3.5 Flash, and it's not slowing down. The company revealed three new AI models today, including its first version of Gemini geared toward cybersecurity. However, none of the new models is the delayed Gemini 3.5 Pro, which was supposed to launch in June.

Gemini 3.5 Flash, which was the star of the show at I/O, has already been deprecated. In its place, developers and users will find Gemini 3.6 Flash. Google makes the usual claims about this model—it's marginally more capable and better at coding, and it has great multimodal features.

Google says the changes to 3.6 Flash were made in response to user feedback on the 3.5 release. In general, Gemini 3.5 Flash didn't appear to live up to Google's promises around code generation. Perhaps that is simply a consequence of Google's intense focus on efficiency as businesses have started to fret over the cost of AI tokens.

Read full article

Comments

Drones capable of spraying water and fire retardants have been practicing how to snuff out wildfires early in California and Alaska this summer—demonstrating a possible rapid-response tool for firefighting as climate change transforms wildfire season into a nearly year-round risk for much of the United States.

Most drones are significantly smaller than crewed aircraft and have shorter flight ranges, meaning they cannot replace large airtankers in delivering massive payloads of water or other fire-suppressing payloads to remote wildfires. But companies and fire agencies—along with organizers of the $11 million XPRIZE competition—are testing whether drones can help firefighting crews respond more quickly to small fires and put them out before they become more destructive.

The California Department of Forestry and Fire Protection, also known as CAL FIRE, ran its own field test involving five autonomous drones that worked together to deploy between 500 and 1,000 gallons of foam combined for suppressing fires on July 15, according to the TV station KPMH. The demonstration was organized with the help of the nonprofit FireWERX and the California-based company Seneca, which is making the drones commercially available starting in 2026.

Read full article

Comments

There are a lot of AI coding applications out there, and as impressive as large language models and the agents they enable have become, many of the most recent developments in AI-assisted development have been in the software that manages those models, not just the models themselves.

Earlier this summer, I spoke with the head of product for Claude Code, Anthropic's Cat Wu, about that company's approach to building that software.

Read full article

Comments

If you’re like me, you or a loved one has struggled through the process of gaining pre-approval for the medical care that your physician has recommended. Personal stories abound regarding the tribulations of patients as they go through hoops to get their health insurer to pay for certain prescription medications, medical procedures, and more.

When used judiciously, this process—known as prior authorization—serves as a check on overuse and spending on services or technologies for which there are less costly alternatives. But a large majority of physicians voice concerns about care delays, which can cause patients to abandon recommended treatments while waiting for the insurance company to verify their eligibility and confirm that the treatment is, indeed, medically necessary. Patients who are denied care may submit an appeal, but that requires more time.

AI might be able to help. With its ability to efficiently sort through vast reams of information, artificial intelligence could theoretically expedite approval of unambiguously allowable claims, thereby reducing care delays. However, AI-driven prior authorization is facing resistance, as it may increase wrongful denials of health insurance coverage. A 2025 American Medical Association survey of physicians revealed significant concern about the application of AI tools, with 61 percent of doctors worrying that AI will exacerbate denials of what they deem are necessary treatments.

Read full article

Comments

As smoke from hundreds of burning wildfires spread across Canada and the United States, the first three operational satellites in the Google-backed FireSat program successfully launched into orbit. The satellites will begin providing wildfire detection capable of spotting even small fires in the United States, Australia, and Europe before the end of the year.

The launch of the microsatellites aboard a SpaceX Falcon 9 rocket from Vandenberg Space Force Base in California on July 7, 2026 marks a transition to “initial operational capability” for the FireSat constellation managed by the nonprofit Earth Fire Alliance. After a three-month testing period, the three satellites will begin actively providing data to fire agencies while covering every fire-prone region on Earth at least twice per day.

FireSat represents the first satellite constellation purpose-built for detecting wildfires, including spotting smaller fires that other satellites may miss. The satellites were designed by California-based satellite manufacturer Muon Space and have received over $15 million from Google to support initial deployment. Other notable financial supporters include the Bezos Earth Fund that committed $26 million.

Read full article

Comments

This week, San Francisco’s attorney general, David Chiu, sent cease-and-desist letters, demanding that Apple and Google remove 13 so-called nudification apps from their app stores, Wired reported.

Nudification apps can make it trivially easy to transform ordinary photos of real people into explicit images. The harmful AI tools allow bad actors to remove clothing, change a person’s features, place them in sexualized positions, and swap victims’ faces onto other people’s naked bodies.

Chiu's letter warned that app stores were violating "California’s laws that prohibit supporting services that create deepfake pornography," Wired reported.

Read full article

Comments

Europe wasted no time using its landmark Digital Markets Act (DMA) to try to rein in Big Tech. Companies like Apple, Meta, and Google have faced steep fines and orders to modify their business practices since the law came into force in 2024. And the hits keep on coming for Big Tech in Europe. After several months of consideration, the European Commission has announced new DMA measures that will force Google to support interoperability and competition in the European Union, and Google is not happy about it.

The new "specification measures" cover two elements of Google's business: Android phones and search. Both changes could theoretically increase competition and give users more choices, but Google claims they will undermine privacy and security. But as a "gatekeeper" under the DMA, Google has no choice but to comply. As the European Commission points out in its announcement, these decisions are legally binding.

On Android, Google will have to open up access to competing AI platforms. Currently, Google's Gemini gets preferential access to the system. Gemini is preloaded on all Google-certified Android phones and can wake up in response to the "Hey Google" hot word. Google's AI also includes system and app automation features, screen content access, and more.

Read full article

Comments

Facing mounting pressure to acknowledge that Grok can still be used to generate non-consensual sexualized images of adults and minors, xAI filed a lawsuit Tuesday, suing the first user that Elon Musk’s firm has accused of using its chatbot to create illegal content.

The complaint targets Terry Wayne Harwood, who was arrested earlier this year for possession and distribution of child sexual abuse materials (CSAM), the South Carolina attorney’s office announced.

As xAI alleged, the company assisted in that arrest after discovering that Harwood had been using two xAI accounts for months to undress or “nudify” non-sexual images of multiple victims, including a young girl who appeared to be as young as 10.

Read full article

Comments

Thousands of unionized Hyundai auto workers began walking off the job early after negotiations with the South Korean automaker broke down over plans to deploy humanoid robots—the most significant pushback from organized labor so far over the latest wave of robotic automation.

The partial strike at Hyundai’s automotive production complex in the city of Ulsan in South Korea represents “the car industry’s first factory stoppage addressing humanoid robots,” according to The Wall Street Journal. Workers have already ended their day and night shifts two hours early at the world’s largest automotive plant from July 13 through July 15, and plan to start staging four-hour strikes from July 20 to 22 after 15 rounds of negotiations failed to reach an agreement, The Korea Times reported.

Union pushback began as soon as Hyundai Motor Group unveiled the latest version of the Atlas humanoid robot, a two-legged robot that stands at more than 6 feet tall and can lift more than 100 pounds, at the start of this year. Atlas is made by Boston Dynamics, the US robotics company that is about to become a wholly owned subsidiary of Hyundai.

Read full article

Comments