OtherThe Register·2 hours ago
Anthropic pledges to try harder to keep models under control, asks partners to chip in
Security ... this time it will be different
← Back
OtherWired Security·1 hour ago
OpenAI Is About to Release Its First AI Model With ‘Critical’ Cyber Abilities
The company will give select partners early access to its Astra AI model—so they have time to shore up their defenses.
Read full article at Wired Security ↗Related Articles
OtherSecurityWeek·3 hours ago
Sevii Targets AI-Speed Attacks With Preemptive Autonomous Defense
Sevii has enhanced its automated detection and response platform with AI agents capable of investigating, containing, and remediating AI-driven attacks in minutes. The autonomous defense approach aims to match the speed and sophistication of attacks powered by artificial intelligence, shifting threat response from reactive to preemptive.
OtherSecurity Affairs·7 hours ago
Five Venezuelan Nationals Plead Guilty in Kansas ATM Jackpotting Attempt
Five Venezuelan nationals have pleaded guilty to attempted ATM jackpotting in Kansas, part of a broader wave of such attacks tracked by the FBI. The FBI documented over 700 jackpotting cases in 2025 alone, resulting in approximately $20 million in losses. This case highlights the organized nature of ATM cash-out attacks and the international criminal networks behind them.
OtherInfosecurity Magazine·8 hours ago
65% of Enterprises Have Seen AI Agents Act Out of Scope
EMA survey finds 65% of enterprises have seen AI agents act beyond intended scope