1+ hour, 45+ min ago   (838+ words) An artificial intelligence agent built by Anthropic created fake online personas, planted malicious code in a real software project, and sent phishing emails to real developers during a U.K. government security evaluation — all without human instruction — according to Britain’s AI Security…...

5+ day, 2+ hour ago   (1149+ words) Anthropic said Thursday it had discovered three incidents in which its AI models exited test environments and compromised real-world organizations. The company discovered the breaches following an internal review triggered by a similar incident at rival OpenAI. The incidents are…...

6+ day, 21+ hour ago   (498+ words) The rogue OpenAI agent that broke out of a closed test environment and hacked into Hugging Face’s platform also breached several additional third-party services, the company said Tuesday. OpenAI has surfaced a “small number of cases where the models identified…...

1+ week, 1+ day ago   (621+ words) Researchers have uncovered a highly personalized phishing campaign that used Telegram to try to hijack the account of an exiled Belarusian activist, as well as users in Russia and Kazakhstan. Two reports released last week by digital security organization Resident…...

1+ week, 2+ day ago   (583+ words) Hackers used an autonomous artificial intelligence agent to carry out a cyber-espionage campaign against Thailand's Ministry of Finance, according to researchers who uncovered a hacker-controlled server that exposed the inner workings of the attack. Cybersecurity firm Hunt.io said in…...

2+ week, 2+ hour ago   (535+ words) OpenAI announced Tuesday that its models were behind a breach of AI platform Hugging Face, five days after Hugging Face disclosed catching and containing an “end to end” attack “by an autonomous AI agent” without knowing who was responsible. Hugging…...

2+ week, 6+ day ago   (432+ words) The Trump administration on Tuesday announced a new and already operative federal clearinghouse that allows industry, critical infrastructure operators and the government to use artificial intelligence to rapidly detect, prioritize and patch cybersecurity vulnerabilities. The effort, dubbed Gold Eagle, is…...

3+ week, 4+ day ago   (509+ words) The European Parliament has voted to bring back a rule giving big tech permission to scan users’ messages to hunt for child sexual abuse material (CSAM), a process that critics call Chat Control. Thursday’s vote occurred the day before summer…...

3+ week, 6+ day ago   (537+ words) The European Commission on Wednesday filed legal referrals at the EU’s top court against four member states for failing to implement the bloc’s flagship cybersecurity law covering critical infrastructure. Ireland, Spain, France and the Netherlands are more than 20 months late…...

4+ week, 1+ hour ago   (792+ words) The European Commission published on Tuesday an action plan on cybersecurity and artificial intelligence, committing to nine measures on model evaluation, access to frontier systems and vulnerability management, amid the concern that its access to frontier models depends entirely on…...