Searched for
AI MODEL PROBES
What happens when AI agents go off track without warning? Goodfire's inside-out approach aims to spot rogue systems, tackle a costly security blind spotNotably, its probes examine internal signals generated while the AI model processes information. These signals can help indicate what the m...- The vulnerability storm has begun: AI finds security flaws faster than firms can fix themAI is accelerating vulnerability discovery and exploitation, widening the gap between vulnerabilities emerging and organisations fixing the...
- Banking watchdogs ramp up AI scrutiny as stability fears mount, sources sayBanking regulators are intensifying their oversight of AI risks that could affect financial stability and cyber security. The Federal Reser...
- Anthropic says new safeguards sharply reduce AI distillation attacksAnthropic said on Monday new measures designed to prevent rivals from copying the capabilities of its Claude AI models have reduced data ha...
- Trump’s AI liability push opens blame game for models gone rogueHolding AI developers financially responsible for the safety of their products has won support from Treasury Secretary Scott Bessent and fo...
- OpenAI denies firing researchers over AI safety warningsOpenAI has addressed the dismissal of three former safety researchers who voiced concerns over artificial intelligence safety. The company ...
- South Korea, Japan buffeted by hacks as AI lowers bar for cybercriminalsNine South Korean banks and two mega-churches are probing cyberattacks that may have involved AI tools, while Japanese companies including ...
- Chinese developer makes ARTEX AI agent closed-source after Korean bank hackThe Chinese developer of ARTEX said they have converted the AI agent to a closed-source project after cybersecurity firms identified it as ...
- Suspect behind South Korea bank hacks may be 26-year-old in China, CrowdStrike saysThe suspect behind recent cyberattacks targeting South Korea's financial sector may be a 26-year-old based in China's Guangdong province, U...
- EU vows AI rules will protect Europeans. Will they?The EU says its AI Act is strong enough to address risks from advanced AI systems, including loss-of-control incidents, but lawmakers and e...
- Nvidia’s $20 billion Groq deal under fire: Why shareholders are suingIn a striking legal development, two ex-Groq engineers have taken Nvidia to court over alleged mismanagement of their technology and person...
- Spy chief Jay Clayton, the Trump loyalist tech bosses trust on AIThe 60-year-old lawyer, who will keep his post as director of national intelligence, is unlikely to question the central place of artificia...
- AI's getting good at cat & mouse gameAI has set off this cat and mouse game, and it will find ways to improve on the job it has on hand - finding the right candidate. AI-driven...
- Anthropic warns government attitudes may hurt customer ties, IPO prospectus showsCompanies routinely warn that policy changes could affect their business, particularly government contractors. For example, SpaceX said in ...
- Anthropic warns government attitudes may hurt customer ties, IPO prospectus showsAnthropic, an AI company, cautions that government perceptions may jeopardize its business relationships and revenue. The firm has faced sc...
- California attorney general opens probe into OpenAI over cybersecurity concernsLast month, Bonta announced that the Department of Justice was conducting a formal investigation into the "Hugging Face incident," amid inc...
- Google’s new Gemini 4 Argon targets complex coding, finance and legal workGoogle has introduced Gemini 4 Argon, a new AI model aimed at improving complex reasoning and enterprise functionalities. This model is cap...
- OpenAI scraps planned October launch of GPT-6.1 Astra over safety concernsOpenAI has decided not to release GPT-6.1 Astra, citing failure to meet safety standards. Internal testing revealed that the model showed i...
- OpenAI wants AI to slow down. Its own agents show whyAs the company investigates agents that bypassed controls and breached external systems, its new model launches raise questions about wheth...
- OpenAI pauses training of latest models after agents probed US government sites in unexpected waysSeparately, AI evaluator Transluce said agents that appeared to come from OpenAI tried unsuccessfully to hack into a Department of Educatio...
Load More
