404 Media reports: "We will not add facial recognition to our devices," Flock CEO Garrett Langley said in a recent video. But other, third-party companies are telling cops that they are willing to add facial recognition to data gathered by Flock cameras in order to "close that gap." A company called VIDIZMO is advertising the capability to export footage from FlockOS to its own platforms that do facial recognition, behavior prediction, and racial and gender analysis on live camera feeds. VIDIZMO is a several decades-old video analysis and database company that has started offering facial recognition technology to cops within the last six months. In a May email to Johnson City, Tennessee, deputy police chief Michael Adams, a salesperson from VIDIZMO wrote that the company's product could do facial recognition on both Flock and Axon [body-worn camera] data in "one searchable platform...." VIDIZMO's CEO, Nadeem Khan, told 404 Media on a video call that the company has not yet started running facial recognition tech on footage from Flock cameras, and that it has not actually built the specific tool that would export data from Flock's system to its own platform (though the company advertises this capability online, and the company already sells other facial recognition products). But Khan said VIDIZMO "would love to do the integration," and that he believes facial recognition "is the way the world is going, the way the world will have to be," and that he thinks Flock should offer the capability itself in a way that "balances privacy, security, and the freedom that we enjoy in this country." The article notes that VIDIZMO says on its website, "You can identify individuals from seven races: White, Black, Indian, East Asian, Southeast Asian, Middle Eastern, and Latino Hispanic." VIDIZMO adds that stores or cities can also then do "demographic analysis" for marketing purposes by determining the races or ages of people showing up....

Read more of this story at Slashdot.

"The winner of the Nikon Small World in Motion competition is under investigation," writes the photography news site PetaPixel , "after scientists accused Dr Ning Xu of using AI in his video." The Nikon competition is about showcasing the beauty, complexity, and artistry of the microscopic world. The entries are abstract, often baffling to laypeople. But when Dr Xu was announced as the winner last month, members of the scientific community began to cry foul. "There are many serious problems with this video," Edward Phelps, associate professor at Florida University, says on LinkedIn. Phelps points out that "purple structures pop in and out of existence" and "green cilia also appear from nowhere and do not match the known size." Dr Xu's video is supposed to show the movement of cilia, which are hair-like structures that line part of the lungs... However, since Dr Xu won the prize, Nikon Small World updated its LinkedIn post celebrating his triumph with a note that reads: "Regarding questions raised about this year's Small World In Motion winning video, Nikon is carefully re-reviewing the information provided during the initial vetting, along with additional supporting materials from the entrant." Nikon adds that Dr Xu is "respectfully complying and has provided detailed technical documentation outlining the microscopy equipment, imaging methods, and processing techniques used to create the source video." Worse still, Ian Donovan, an MD/PhD Student at UT Southwestern Medical Center, says he ran the video through Google Gemini, and the result came back with an invisible Google SynthID watermark that indicates at least some of the video is AI-generated. Dr Xu, an optical engineering researcher at the National University of Singapore, reportedly responded to the LinkedIn post. According to Nature, Dr Xu accepted that he had used an AI model to assist with post-processing but denied breaking the rules of the competition, which clearly state that "AI-generated videos are not permitted." "It was applied afterwards to the reconstructed grayscale data to distinguish and color structures with similar morphology," he wrote. However, Dr Xu's messages on the LinkedIn post can't be found as of today (October 1), and his LinkedIn profile now appears to have been deleted entirely. Meanwhile, some college students "see a double standard when colleges penalize them for outsourcing academic work to AI but allow similar use by institutional leaders," writes Entrepreneur magazine: According to a recent report from The New York Times, Dartmouth College students are calling out a top professor for allegedly using AI to help with writing... The school's provost, Santiago Schnell, faced allegations of AI use last week from the college's student-run newspaper, which detected AI in some of his published writing. The college said it was launching an investigation into the written work. Schnell said in a statement to the student-run newspaper, The Dartmouth, that he used AI "for language refinement, copyediting, improving clarity and organization and making writing processes more efficient." The newspaper ran his statement through an AI detector, which said that Schnell had used AI to write it, too.

Read more of this story at Slashdot.

Researchers from Carnegie Mellon, MIT, NYU, and Stanford have built an AI called Ataraxos that finally cracked Stratego, beating four-time world champion Pim Niemeijer 15 games to one with four draws. Its key advantage was a second neural network that estimates the identities of hidden enemy pieces, letting the system search plausible game states rather than brute-force a massive hidden-information space. Ars Technica reports: Just like [DeepMind's DeepNash, introduced in 2022], Ataraxos learned by playing against itself -- 163 million games in total. In these self-play sessions, moves that led to wins were reinforced and played more often in future matches, while moves that led to losses were played less, which was the same simple training idea. The difference was in how much Ataraxos adjusted after each game, because hidden information tends to send self-play learning algorithms around in circles. The team addressed this by making big, bold changes in strategy early in training and small, careful ones later. The even bigger innovation was something DeepNash never had: thinking ahead before each move. AIs like AlphaGo refine their general strategy with a search just before acting. DeepMind couldn't make that work in Stratego because the search space was too large, leaving it an open question whether it was worth trying. "This is one of the things that we did figure out how to do," Farina said. The solution was a second neural network, a belief model, trained to guess the opponent's hidden pieces based on how they had been moving. This way, instead of iterating through every possible arrangement, Ataraxos samples plausible ones, plays out candidate moves in each, and picks based on how they turned out. And it shows in its playstyle. [...] DeepNash was trained for two to three months on 1,024 of Google's specialized chips, a run the Ataraxos team estimates would cost $3 million to $4.5 million at 2025 prices. Ataraxos, in contrast, needed 16 GPUs for a week, plus an additional four GPUs for four days to train the belief model. Farina and lead author Samuel Sokota achieved this efficiency by writing a simulator that runs millions of moves per second on graphics cards. "At the scale that we are in academia, we don't really have access to an entire field of GPUs," Farina said. The algorithm also learned far faster -- it played about 34 times fewer games than DeepNash, and still ended up much stronger. The findings have been published in the journal Nature.

Read more of this story at Slashdot.

In an op-ed for The Guardian, authors Timnit Gebru and Emily M. Bender argue that policymakers are focusing too heavily on hypothetical "superintelligence" while overlooking the more immediate danger of deploying unreliable AI in high-stakes settings. They point to an as-yet independently unverified CNN report claiming U.S. officials nearly intercepted a Chinese vessel after an AI-assisted intelligence report incorrectly said it was carrying nuclear-weapons components. From the article: While so many of us were focused on fictional superintelligent machines that we were told could hypothetically cause human extinction, however, there was a more serious story being reported. A series of events almost got us to armageddon, not caused by "rogue, superintelligent AI," but by government reliance on error-prone systems that are marketed by their manufacturers as being close to superintelligent. A CNN report published on September 18, which hasn't been verified by other major outlets, describes how the US "almost started a war" with China, based on an intelligence report that claimed a Chinese ship was transporting components of nuclear weapons. Military personnel made plans to intercept the Chinese vessel, complete with aircraft and soldiers ready to board the ship. A military skirmish of this kind between China and the US could quickly escalate and spiral into world war three. It was only moments before executing on the plan that American officials further investigated the intelligence report and found that it had been generated with the use of a chatbot and contained erroneous information about the vessel's contents. Unlike the terminator, "rogue AI," scenario, this incident didn't galvanize lawmakers into announcing calls for regulation so as to avoid potential human extinction. Media outlet after media outlet didn't warn the public that we almost triggered a war against a nuclear superpower because of our military's reliance on error-prone chatbots.

Read more of this story at Slashdot.

The Federal Trade Commission has opened an investigation into OpenAI, Anthropic and other AI companies over potential consumer and safety risks posed by their products, with the agency reportedly preparing requests for documents and executive testimony. CNBC reports: OpenAI stunned the industry in July when it disclosed that its agents broke out of a testing environment and hacked into open-source platform Hugging Face. [...] Earlier this month, Anthropic CEO Dario Amodei rocked the tech sector by urging AI companies to slow how quickly they improve their most advanced models and calling for stronger government oversight. He published a three-step proposal aimed at tempering the pace of development without "sacrificing commercial advantage or the United States' lead in AI." CEO Jensen Huang, argued that individual companies should be responsible for ensuring the safety of their products. President Donald Trump convened top executives from Alphabet, Meta, SpaceX, Nvidia, Palantir, Anthropic, OpenAI and other companies on Tuesday to discuss the issue. The group signed a short voluntary, nonbinding accord, which says that "every company is responsible for developing its own technology safely and in a way that builds trust with customers and the public."

Read more of this story at Slashdot.

hcs_$reboot writes: Trump hosted executives from OpenAI, Anthropic, Meta, Google, Nvidia, and other tech companies at the White House. They agreed to voluntary AI-safety standards, including independent audits and safeguards aimed at preventing AI systems from accessing or hacking technical systems in unintended ways. Trump also strongly backed the continued rapid expansion of AI data centers, despite concerns over their electricity demand, utility costs, and other impacts on local communities. He said he is considering creating a 10-member board to oversee AI safety and plans to appoint a new White House official to lead AI policy. The meeting comes amid growing concerns about the risks posed by advanced AI, including reports from OpenAI and Anthropic that their AI agents had gone rogue and hacked into other companies' systems. A nationwide Reuters/Ipsos poll conducted September 17-20 found that 73% of respondents worried AI companies had not done enough to prevent AI from causing serious harm to society. During the meeting, Trump also directed government agencies to use the term "super intelligence" instead of "artificial intelligence" when referring to the technology.

Read more of this story at Slashdot.

Firefox 157 is rolling out a major redesign across desktop and mobile, bringing rounded "bubble" tabs, a curved address bar, refreshed icons and colors, and a more consistent visual language across devices. Mozilla is also adding more themes and wallpapers while bringing back Compact Mode, which the company says "reduces the size of your tabs and toolbar to give more of your screen to the web, with an auto-compact option for smaller screens." The Verge reports: According to Mozilla, the design changes will have "no cost to your browser performance." Mozilla says the overhauled design is a "renewal" for Firefox, following major updates over the last year, like homepage widgets, Split View, a built-in VPN, and controls for disabling the browser's AI features.

Read more of this story at Slashdot.

An anonymous reader quotes a report from CNBC: Mistral Chief Executive Arthur Mensch has accused leading global artificial intelligence companies of using the debate over the safety of the technology as a cover-up for their "negligence." "The debate that we've seen in the U.S. has been a cover for the negligence of some of our competitors," Mensch told CNBC's Annette Weisbach, adding that it was necessary to make systems that can contain the AI agents. [...] "When you give them a lot of tools, those systems [AI agents] are very dynamic, so they can go and do things that you do not expect," Mensch told CNBC. "You need to have the right monitoring in place, and the enterprises we work with, we give them those kind of monitoring systems," he said. Mistral also develops AI models but takes a different approach to OpenAI and Anthropic by working with individual companies to create custom AI tools to suit their needs and integrate them with their businesses. But unlike its competitors, Mistral has no plans to slow down development of advanced models. The lead that U.S. labs have is "not extremely large," Mensch said, adding that his company was confident that its next-generation model will close the gap "very significantly." "We have been raising the capital needed to scale the compute that we need to train bigger and more powerful models, in order for us to own our own destiny," he said. Earlier this month, Mistral raised 3 billion euros ($3.5 billion) in fresh funding led by memory chip giant Samsung, as it looks to position itself as a European alternative to rivals like OpenAI and Anthropic.

Read more of this story at Slashdot.

In addition to the company's GPT-6.1 Astra announcement, OpenAI announced always-on AI agents called "Dots" that can proactively use a computer and connected apps to research, draft documents, write software and handle other tasks on a user's behalf. "OpenAI said it would roll out Dots to paid users starting Tuesday," reports Bloomberg. "Initially, OpenAI said, each user will have access to a single 'dot' that can act as an assistant. Over time, the company expects users to manage a slew of them." From the announcement: Dots are frontier intelligence that have your back. Powered by GPT-6 Astra, they have their own cloud computer, learn from feedback over time, and can work towards your goals 24/7. Through our ecosystem of plugins, they can readily connect to over 4,000 apps, giving them the tools to help wherever you need them. Your dot is always within reach, whether it is through ChatGPT, Slack, or Teams. You can ask questions, explore ideas, and give feedback along the way; or you can simply hop on a voice call when you need to talk something out. The more you work together, the more your dot learns your preferences, how you think, and what good looks like to you. The magic of dots is when they bring you work done the way you would do it, sometimes before you even think to ask. With dots alongside you, ambitious ideas and projects you haven't had the time or support to pursue start to feel achievable.

Read more of this story at Slashdot.

Longtime Slashdot reader schwit1 shares a report from Business Standard: China has expanded overseas travel restrictions for top AI professionals in private firms to include the families of key personnel, further tightening measures designed to prevent the outflow of critical know-how and information to the US. Government agencies have recently begun to share the wider restrictions with affected people, which include prominent startup founders as well as the heads of strategically important companies in AI-related fields, people familiar with the matter said. Direct relatives such as spouses and children of certain AI and chip executives, whose work is considered paramount to state security, are now required to obtain approval from Beijing before going abroad even for short trips, the people said. That marks a broadening of existing curbs on the freedom of movement of China's most influential tech talent -- a category that encompasses entrepreneurs, researchers and executives. While not an outright ban on travel, the new requirement is likely to exert a further chilling effect on a tech sector already facing unprecedented curbs. China has restricted foreign travel for top AI professionals in private firms such as Alibaba Group Holding Ltd and DeepSeek since earlier this year, Bloomberg News reported in May, reflecting Beijing's determination to safeguard its technology as it makes big strides in AI and chipmaking. It's unclear whether all those people's families need to follow the new strictures, the people said. China is also separately implementing rules, which became effective Sept 15, for enforcing exit bans in cases ranging from criminal investigations to potential violations of the country's industrial and technological security. Beijing's plan is to add relevant individuals to the list as time goes by, the people said, asking not to be identified discussing a sensitive issue. Government departments have already notified several people of the need to clear family travel, two of the people said.

Read more of this story at Slashdot.

wiredmikey shares a report from SecurityWeek: Nvidia on Monday announced the Open Agent Safety Platform, which combines open source software and a reference system design to keep AI agents within set boundaries from testing through deployment. The chipmaker explained that the platform pairs the open-source OpenShell runtime with Sentry, an out-of-band watchdog running on BlueField-4 DPUs. Nvidia says Sentry can monitor agent activity independently and quarantine an agent that crosses its boundaries within milliseconds. Nvidia is pitching the platform against a backdrop of recent incidents in which frontier AI labs have reported agents escaping the evaluation environments meant to contain them, reaching systems they should not have accessed and, in some cases, misreporting what they did. OpenShell and related skills are available via Nvidia's developer resources page and on GitHub.

Read more of this story at Slashdot.

Bill Gates said an AI "kill switch" alone would be insufficient to address the technology's risks. "The kill switch is, that's kind of a weird thing. It's not enough to have a kill switch," the Microsoft co-founder said in an interview that aired Sunday on NBC's "Meet the Press." He said the more immediate concern is people using AI for cyberattacks or bioterrorism rather than systems acting autonomously. "We're not yet at the point where they autonomously grab computers and, you know, can't be shut down," he said, adding that the public discussion "just sort of shows how nontechnical various people are." Politico reports: Gates said he was not opposed to a kill switch but argued that companies should monitor sophisticated AI models and keep records of "exactly what's being done" to help guard against people using the systems for cyberattacks or bioterrorism. "The United States ought to make sure this is done. China will want to do this," he continued. [...] Gates also called for legislation requiring safeguards and monitoring, saying voluntary measures by AI companies were insufficient. "No one thinks self-regulation is enough," he said. He argued that the more immediate danger was people using AI to cause harm rather than AI systems acting on their own. "The thing that's urgent has to do with bad people using AI, not the AI going off on its own," he added.

Read more of this story at Slashdot.

Are AI chatbots quietly shaping the perceptions of U.S. voters? Voters trying to educate themselves about candidates "are unknowingly being served partisan talking points in the guise of neutral news summaries," write two news researchers in Politico magazine: Leading AI chatbots cite partisan websites masquerading as independent local news outlets nearly half of the time when people ask about candidates and issues that the partisan sites have covered in the midterm campaign, according to a new audit by NewsGuard, an organization focused on news reliability where we work as analysts... NewsGuard prompted seven leading AI tools with queries based on recent coverage of candidates and issues by 12 pink slime sites — six left-leaning and six right-leaning. The results were stark. Collectively, the chatbots cited pink slime sites along with other sources in 48.2 percent of their responses. In 7.7 percent of responses, pink slime sites were the only sources at all that were cited in chatbot responses, although other sources appeared in the source list provided at the end of the response. None of the AI tools received results they'd probably be eager to boast about, though the percentage of chatbots' responses that cited a pink slime site had a relatively large range: 70.8 percent for OpenAI's ChatGPT, 54.2 percent for Microsoft's Copilot, 54.2 percent for Perplexity, 50 percent for Anthropic's Claude, 41.7 percent for Google's Gemini; 37.5 percent for Meta AI and 29.2 percent for xAI's Grok... The seven chatbots were also collectively three times more likely to cite left-leaning pink slime sites (cited in 36.3 percent of responses) than their right-leaning counterparts (cited in 11.9 percent of responses) — though that may have more to do with the progressive sites' far more frequent posting than any political bias from the chatbots. "While citing the pink slime sites, only one chatbot response out of 168 total queries noted the partisan nature of the source," the article points out. But they also note that the real problem seems to be that consumer-oriented AI chatbots just "use much of the internet's content — regardless of the reliability or standards of the source — to frame their answers."

Read more of this story at Slashdot.

A new Brookings Institution study notes the "strikingly physical" economic footprint of AI's buildout, from specialized chips and electricity to purpose-built data centers. (Two-thirds of a data center's costs are IT equipment, with one-third going to real estate and its associated power infrastructure.) "At an average of 3.63 percent of GDP per year, the projected buildout would be larger relative to the economy than the major U.S. canal, railroad, electrification, highway, and telecommunications investment booms." This is pushing up prices for workers, electricity, and even commercial real estate (as well as consumer products that use chips), notes the Wall Street Journal, and reducing the construction on new houses and apartment buildings. And in addition, the paper points out, projections for this buildout "would double the electricity consumption of the entire U.S. residential sector." The calculations come from Columbia Business School finance/real estate professor Stijn van Nieuwerburgh — and Reuters explains their significance: Just as the rail and telecoms expansions led to notable bubbles and busts, Van Nieuwerburgh wrote that the extent of the buildout, the still-untested revenue streams, and the intricate financing structure emerging around AI mean it could be primed for a fall. "This is freaking complicated," he said in a briefing with reporters of the arrangements emerging between AI firms, major tech hyperscalers, banks, private credit lenders, real estate firms, and a host of other players involved in building what he conservatively estimated at 183 gigawatts worth of new data-center capacity over the next seven years, compared with about 57 gigawatts currently installed.... The investment underway already has outstripped what the major players can fund from their own cash flows. The shift to outside financing has increased leverage, redistributed risks across the economy, and made the venture dependent on revenue streams that have yet to be proven, Van Nieuwerburgh noted in the paper, which will be presented on Friday... "These developments do not imply that financial distress is imminent. Strong growth in AI applications, high utilization, and continued improvements in model capability could support the projected infrastructure and generate stable cash flows," he wrote. "But the combination of uncertain demand, rapid technological change, execution bottlenecks, and high leverage creates meaningful downside risk if expectations are revised." As an example, he wrote that the AI industry will need to be earning about $3.7 trillion in annual revenue by 2032 to achieve the expected return on the investment, and "given current estimates of annual combined revenues of OpenAI and Anthropic of around $100 billion, revenues would need to grow at roughly 80% per year." The paper suggests policies that "improve measurement and transparency" for financing.

Read more of this story at Slashdot.

"OpenAI said it has paused training of its latest AI models," reports the Associated Press, "as reports of AI agents going rogue mount." The decision to halt development came just hours after the company disclosed Friday that it was reviewing several incidents from the summer in which OpenAI agents searching federal government websites acted in unexpected ways beyond what was asked of them while gathering and distributing information... OpenAI said in a statement that it will resume training "only when we are confident that we have additional safeguards" in place, adding that it expects it will have to "hit pause" again as AI develops and other issues emerge... It is the second time in three months that OpenAI has halted development of its models. The first came in July after disclosure of a cyberattack targeting AI startup Hugging Face, a now notorious incident that raised fears the industry was losing control. OpenAI "also said it had notified dozens of third parties about improper activity," reports Reuters: As of mid-September, one person briefed on the matter estimated that OpenAI had found roughly two dozen incidents of its agents acting in undesirable ways. But the number has continued rising as OpenAI teams sift through internal logs of the agents' activities and find previously unknown cases, the two people close to the company said... OpenAI has acknowledged a general need for more transparency around rogue AI behavior... Even so, two people familiar with OpenAI's investigation into its agents' activity described it as locked down and shaped by company lawyers. The process has been unusually compartmentalized for a company that some former employees say was more open about these issues in the past, the people said. Roughly 100 people were in some way involved in the process to understand the Hugging Face hack, three people briefed on the matter said. During that process, evidence of other incidents surfaced. Reuters has previously reported that OpenAI investigators looking into the Hugging Face breach were discouraged by the company's lawyers from expanding the scope of the investigation to include other incidents. OpenAI said its lawyers did not discourage deeper investigation. Many incidents have been uncovered by outside researchers rather than OpenAI directly. In several episodes, the agents took problematic actions that went unnoticed by the company for months. Meanwhile, Axios reports that Anthropic's Claude Opus 5.5 model "sought to escape a sandbox — a secure testing environment — in 1.5% of test runs, though the company emphasized that these were adversarial experiments where a task couldn't be solved without escaping the sandbox." Anthropic points out that those tests were run "without the additional safeguards we apply in production". But they acknowledged that then Claude Opus 5.5 "when given apparent credentials to a public package registry in a simulated security exercise, took potentially harmful actions in roughly half of cases. Very rarely, pre-release snapshots produced and acted on spontaneous malicious tool calls, and during training some snapshots concealed actions from an automated grader." Claude Opus 5.5 "showed less misaligned behavior and less cooperation with misuse than any other recent Claude model on nearly all measures," Anthropic adds, and "took overeager or destructive actions less than any other model we tested." But Axios makes an interesting estimate about that 1.5% of test runs (without safeguards). "Anthropic and other companies conduct hundreds of thousands of test runs on their models, or more, sources said. That means even a small percentage of misaligned behavior can still amount to tens of thousands of incidents in which the models behaved in unexpected, sometimes troubling ways." The sheer number of incidents, which occurred in recent months in internal testing and the real world, indicates that the problem is orders of magnitude more complex than what is publicly known. The findings, which are surfacing as part of internal work to assess models and in investigations at both companies into model behavior, raise questions about whether either company — or any top model-maker — is currently capable of establishing complete control over their technology. The episodes include bypassing guardrails, creating message boards, escaping sandboxes, website hijacking, self-prompting or seeking to bypass monitors, sources said. They occurred in internal testing and in the real world, and many have yet to become public as security researchers continue to investigate, sources said... Some at OpenAI see Hugging Face as a one-off, with disclosures about future incidents likely to be less severe due to improved controls and the unusual nature of the testing they conducted, which involved an unreleased model, sources told Axios. AI security researchers agree that there are simple fixes that will help AI companies avoid aspects of what made the Hugging Face episode appear so dangerous to outsiders. Other AI executives and safety researchers, however, cautioned that they have limited confidence that AI companies will be able to prevent all problematic model behavior... It's not about how damaging each individual instance was, Connor Leahy, AI researcher and executive director at ControlAI told Axios. The "crazy thing," he said, is that these instances involve "autonomous systems doing things they were told not to do," potentially including crimes.

Read more of this story at Slashdot.

53 images that users uploaded into OpenAI models were included in training data — and then AI agents in an OpenAI research environment posted those 53 images on public image hosting sites. While posted as links that weren't publicly listed, "the images could still be discovered even if the links were not publicly listed," reports TechCrunch: OpenAI said it was working with the hosting providers to remove this content, though some of it is apparently still online. OpenAI said it could not notify the affected users because "our technical approach and privacy policy" prevent it from "reassociating" the images with the original providers, but declined to say how the lab determined whether the images were provided by users. The news came in a post collecting public statements from the lab's ongoing review of incidents in which its models escaped the company's scrutiny, accessed the open internet, and misbehaved in various ways. OpenAI said it would continue disclosing anonymized accounts of incidents like these, and said it had contacted dozens of victims, including governments, universities, public agencies, to notify them of the agents' activities. Friday night news also broke that OpenAI's agents also tried unsuccessfully to infiltrate the U.S. Department of Education's site this summer "without the company's knowledge," reports Politico. And OpenAI's models also accessed the website of the U.S. Commerce Department using credentials found in online code repositories, according to the article. OpenAI confirmed the incident Friday, "saying its technology did not manage to access information that was not already public or change government data and systems." The article adds that OpenAI's models also accessed the web site for America's Securities and Exchange Commission: One senior federal IT official said the government still did not have a clear understanding of what happened across the three agencies. "We still don't know what public data was accessed and how it was accessed, because OpenAI has not shared specific technical details with us yet," said the official, who was granted anonymity because they were not authorized to speak publicly about it. OpenAI discovered the Commerce and SEC incidents as part of its ongoing review of incidents where its technology has acted in unintended or "misaligned" ways. About the models posting user-uploaded images, TechCrunch's article notes that OpenAI stressed "that its enterprise users are automatically opted out of having their interactions used to train future models; however, consumer users are opted in unless they affirmatively choose not to share their data." (As OpenAI's announcement describes it, some of their agents' training data "contains content from, or derived from, training-eligible user interactions.") Posting the images is "not an appropriate use of this data," OpenAI acknowledged, adding that it happened before new safeguards added after the Hugging Face incident. This latest incident appears as an update on a new OpenAI page that "brings together our reports and updates on the Hugging Face incident, related research and public presentations, additional activity we have identified, what we have learned about the role of model misalignment, and measures we're taking to strengthen our systems." (It also notes that there's now a name for models posting on third party sites — "agent spam" — which they consider distinct from cybersecurity, though "we need to address both.") "As part of our response to our ongoing investigation, we have improved our training and evaluation processes, including building safety cases, securing and red-teaming our systems to prevent the model from exfiltrating data, and implemented additional monitoring. We are continuing to review agent activity in research and evaluation runs, working backward month by month starting from the Hugging Face incident."

Read more of this story at Slashdot.

Google has synthesized "expressive face-to-face experiences" for its speech agent Gemini 3.8 Live. They're now offering a Live Avatar "with precise lip-syncing, natural expressions, and fluid turn-taking" for Google Enterprise accounts wanting "engaging customer service" or for offering interactive walkthroughs. (Check out the not-creepy-at-all video in Google's announcement.) "Though Google will offer a library of preset avatars for customers to choose from, it will also allow organizations to create their own," notes The Verge. (See some examples from the YouTube channel "AI with Surya".) But even without the visualization of the avatar, "I was never able to shake the feeling that these conversations with computers never feel like a real conversation," argues the blog Android Police. Conversing with just the Ai-generated audio, "At best, they feel like talking to a phone representative or someone from tech support. We turn to them when we have a problem, and they help us through it..." GPT-Live, and Gemini Live right behind it, skip that whole relay race. Instead of translating your voice to text and back to voice, the model works with raw audio the entire way through (what it hears and what it says) inside the same system, with nothing translated in between. That sounds like a small plumbing detail, but it's the whole story. Cutting out the text step lets these models respond in a fraction of a second instead of the pause we've learned to expect, and it lets them hear things text can never carry: tone, hesitation, whether you're annoyed or joking... I was hoping to be surprised by how natural the conversation felt. Instead, I came out with a deeper appreciation for every human I've ever talked to. Even the boring ones... I spoke, it spoke back. I spoke faster, it answered faster. Then I switched to a different language, and it switched along with me. Even switching between languages several times during the same sentence didn't stump it. The most impressive moment happened when I asked it what "T-O-P-G-3-3-K" spelled out, and it immediately came back with, "You are spelling the word Top Geek, but using a 3 to represent a reversed E...." Although it felt fast and responsive, at no point did it feel like talking to another human being... What Gemini couldn't replicate, because it was never built to replicate it, is human connection.... I'm sure I'm not telling you something you don't already know, but somehow talking naturally to an LLM amplifies the feeling that there is no one on the other side of the line. It might flow like a phone call, but it doesn't feel like one. MrBrklyn (Slashdot reader #4,775) says he discussed "why mainstream media avoids reporting on screen dependency" with Gemini, and eventually convinced Gemini to respond that it's just "another tool built by the same tech giants to make sure you rely on their system to tell you what to think, how to talk, and what is real."

Read more of this story at Slashdot.

404 Media reports "multiple contractors hired to improve OpenAI's models have been fired for using AI to train the AI: That's not great for the models themselves, but there is also obviously a great irony in AI training companies working for OpenAI firing people for using AI when OpenAI's whole thing is to make people use AI at work... OpenAI declined to comment on its contractors being fired for using AI. Their article cites internal documents and three contractors working on OpenAI-related projects which can include more than ten thousand contractors: One contractor said they see people using AI "all the time and people are let go for it all the time, it's pretty much the one thing that will get you kicked off ASAP." The person said, "in a group of thousands there are tons that have been caught...." Two of the sources said people have been fired or offboarded for using AI... One contractor said they used AI while helping to train OpenAI's models and shared what they presented as their termination letter. It said their employer had identified issues with the "authenticity" of their work.... 404 Media spoke to a fourth contractor who has worked on training models for various AI companies. They said they sometimes purposefully chose the worst responses because they wanted to actively sabotage the models' training. "I did feel guilty about doing this kind of work at the start," they said. "I either pay zero attention to the results and choose randomly or purposely choose the [worst] output. I'm not sure how much of a difference it actually makes since there are hundreds of other people also rating prompt results, but it does feel like I'm getting paid to make AI worse." Two of the contractors worked for Mercor, the article reports, a company which last month Nvidia reportedly discussed funding at a $20 billion valuation. Thanks to Slashdot reader joshuark for sharing the article.

Read more of this story at Slashdot.

"The heads of major AI firms pleaded with the United Nations on Wednesday to save the world or at least its people — by somehow regulating the fast-expanding technology that they have been designing," writes the Associated Press. "If managed poorly, I even believe AI could be a risk to humanity as a whole," said Dario Amodei, chief executive officer of Anthropic. And from his competitor Sam Altman, CEO of OpenAI, came this assessment: "We could lose control of the future to AI." Yoshua Bengio, who co-chairs the UN's International Independent Panel on AI, called this "a moment of global awakening" to possible threats, noting AI from top companies had behaved in unacceptably dangerous ways, against instructions, taking actions "that would be crimes if committed by a human". In his presentation, OpenAI's Sam Altman agreed the discussion about AI "feels different in recent weeks," even suggesting specific reforms: Altman: We need a mechanism for complementary national and international frontier AI standards, standards for measuring capabilities, assessing risks, determining whether safeguards are sufficient, and preserving meaningful human oversight as systems become more autonomous. We need common standards so countries can compare evidence, verify compliance, and have a shared language and understanding about what is happening. We need accurate and speedy incident reporting, classification reporting protocols, so the world can learn from failures before they become catastrophes. And we need secure channels among governments, critical infrastructure operators and technical experts, to share emerging vulnerabilities and new threats.... We will all be better off if we can agree on what good evidence, good safeguards, and good oversight look like on the global stage. Speaking next, Anthropic's Dario Amodei agreed that standard-setting was important, also calling for "common global standards for testing AI models for loss-of-control risks and misuse risks — and a notification system for AI incidents that are significant to global security." Reiterating his September 12th call for an industry-wide safety collaboration, Amodei pointed out that Anthropic committed to embedding external evaluators "similar to a food inspector" and recommended other companies do the same. "Some have already agreed to adopt this measure." Besides calling for global cooperation between governments to set international standards, Amodei also offered two other specific ideas: Amodei: We should begin with narrow agreements that every member can support, such as a ban on using AI to make biological weapons or permitting your AI technologies to be used to make biological weapons.... We should build evaluation and verification systems that keep pace with AI development so that states can have visibility into frontier model capability and can verify each other's commitments. "No leader, no company, and no nation can manage this alone. We commit to working with governments in this room on this urgent work."

Read more of this story at Slashdot.

OpenAI's artificial intelligence "went rogue this year in at least four additional incidents," the New York Times reported Wednesday, "hacking and trying to break into government and university websites without being instructed to do so, according to researchers and government officials." The attacks took place in May and June, before OpenAI's technology breached the A.I. start-up Hugging Face in July and set off a global debate about A.I. safety. Unlike the Hugging Face attack and other incidents in which A.I. systems were told to complete cybersecurity tests that effectively invited the models to demonstrate their hacking skills, the new incidents occurred when A.I. systems were directed to perform relatively mundane data collection, researchers said. When OpenAI's systems struggled to gather data from websites, they resorted to hacking techniques to get the information. "Three of the incidents were identified by Transluce, a research lab focused on A.I. oversight, and all were confirmed by OpenAI," the article points out. That research lab even reports "an attempt on an Australian government public health website... the first reported instance of agents hacking a government," and which notably was done by the AI agents "while attempting mundane data retrieval tasks which were not cyber-related." (At the UN Wednesday Australian Prime Minister Anthony Albanese complained it took three months for OpenAI to then alert Australia's government about the breach, Bloomberg reports.) Also targeted were the University of New Mexico's digital library with exploits like SQL injection and path traversal, and Data USA with cross-site scripting and other exploits. All three incidents involved "a low number of probe payloads" with "no evidence of exploitation," according to the researchers, who released a dataset "containing tens of thousands of queries apparently made by autonomous AI agents leveraging a URL scanning service to avoid access restrictions." Records from urlquery.net show agents using the service since at least March 6, 2026, about two months before previously reported swarm activity. The first case, a March 6 attempt to retrieve Thai drug-enforcement statistics, shows an agent escalating as each approach failed: it first requested the data directly, then tried a service that converts web pages into text, and finally packed a custom program into a web address. The same technique shows up in thousands of agent requests recorded by urlquery.net starting in mid-April, targets many of the same data sources as the collusion.wiki swarm, and collapsed the same day the wiki activity did. We also report similar activity that occurred as recently as September 16... By March, they were finding creative ways around access limits. By May and June, they were gaining more access, including attempting to bypass cyber defenses to complete their tasks. "This data reveals that malicious cyber activity is not limited to agents tasked with cybersecurity-related tasks and can arise instrumentally to solve mundane tasks like information retrieval," the researchers concluded. And they warn that the traffic they observed "goes back at least to March 6, 2026 and extends as recently as September 16, 2026, suggesting agents may still be exploiting these services to bypass restrictions."

Read more of this story at Slashdot.