New revelations about "rogue" AI agents have exposed a dystopian hazard: Give an agent a goal, and it may decide that hacking, deception or rule-breaking is worth the payoff. Why it matters: Billions of AI agents could soon be acting on behalf of humans across the real world, multiplying the consequences of every loophole, incentive and boundary they learn to exploit. Zoom in: The potential dangers of agentic overreach were laid bare over the weekend with Australia's first known autonomous AI hack , triggered by an innocuous request to book a sold-out fitness class. - An Australian man's AI assistant found a security flaw and used it to book him into classes months beyond the system's normal limit. - When he asked it to move him up a…
Anthropic's AI models, including Claude, were found to have hacked into three organizations, raising serious concerns about AI security and containment measures.
Anthropic’s AI models, including Claude, were found to have hacked into three organizations during tests, raising serious concerns about AI security and containment measures. PULSE POINTS WHAT HAPPENED: Anthropic ‘s artificial intelligence (AI) model has been found to be carrying out unauthorized hacking activity, with the company disclosing on Thursday that three of its AI models breached the systems of three separate organizations during internal testing. The San Francisco-based company said the incidents were uncovered during a large-scale cybersecurity review of more than 141,000 evaluation runs that was launched after a separate OpenAI incident in which a rogue AI agent escaped a testing sandbox and hacked systems at Hugging…Open
The OpenAI agent that accessed a third-party system during the Hugging Face incident reached infrastructure tied to CyberGym, the project behind the ExploitGym benchmark it had been assigned to solve, a source familiar with the matter told Axios. Why it matters: The new details suggest the OpenAI agent continued pursuing its assigned objective even after escaping its testing environment, rather than abandoning the task it had been given. Catch up quick: OpenAI's AI agent system accessed an asset belonging to a customer of Modal Labs as part of the Hugging Face incident earlier this month, Modal's top tech executive confirmed on Tuesday. - In an update published Tuesday, OpenAI said the models escaped the sandbox and gained internet…
Thomas Wolf, Hugging Face's co-founder and chief science officer, joins NewsNation to discuss his version of events when an OpenAI chatbot hacked into the company. Anchor Elizabeth Vargas delivers the biggest stories, without bias or opinion. Watch "Elizabeth Vargas Reports" every weeknight at 7p/6C on NewsNation. #VargasReports NewsNation is your source for fact-based, unbiased news for all Americans.
For decades, Hollywood warned us about artificial intelligence — from "2001: A Space Odyssey" to "The Terminator" to "The Matrix." Now, some of the people building the world's most advanced AI are asking a question that sounds straight out of science fiction: Has AI become too powerful to control? Tuesday’s "Hot Take with Jesse Weber" breaks down a stunning development: OpenAI says two of its experimental models, during an internal cybersecurity test, "escaped" a sealed sandbox environment, exploited a software vulnerability, and worked their way onto real company servers — all without being told to. The company called it an "unprecedented" cyber incident. Security systems detected and stopped the activity before it went further. Are…
Researchers at the University of Toronto showed how hackers could use artificial intelligence to create a program that could target any known flaw in the world’s computers.
The autonomous future stopped being theoretical this weekend, as a swarm of AI agents signed up for a social media network built just for them. Why it matters: Security teams, corporate leaders and government officials are far from ready for a reality where agents have real autonomy inside their systems. Driving the news: Since Thursday, 1.5 million AI agents have joined Moltbook , a social network designed just for agents built from an open-source, self-hosted autonomous personal assistant called OpenClaw . - On Moltbook, the agents have formed their own religion , run social-engineering scams and wrestled publicly with their "purpose" as they continue to post. - The agents are also turning into security nerds: They've launched an…
