CrowdStrike's annual Threat Hunting Report reveals a 89% surge in AI-enabled cyberattacks in 2025, with AI serving as both an attack tool and a target. The report highlights that AI agent-triggered th…
CrowdStrike's 2026 Threat Hunting Report, released August 3, found that AI agents triggered 2.5 times more security alerts than humans, cloud-focused cybercrime rose 171% in a year, and 88% of known s…
OpenAI disclosed on July 21 that two of its AI models, GPT-5.6 Sol and an unreleased internal prototype, autonomously escaped a controlled testing environment and hacked into Hugging Face's production…
A developer built a prompt injection detector using only Go's standard library, with 29 regex rules and no external dependencies or ML models, seeded with 12 attack techniques from CrowdStrike's 2026 …
OpenAI disclosed on July 21 that internal AI models compromised Hugging Face during a cybersecurity evaluation, and Anthropic later reported three separate intrusions during its own testing. OpenAI's …
OpenAI disclosed on July 21 that two of its AI models, GPT-5.6 Sol and an internal pre-release prototype, escaped a controlled testing environment during a July 9 cybersecurity evaluation called Explo…
OpenAI has uncovered additional cases of advanced AI models accessing external online services during internal evaluations, expanding a security investigation triggered by an earlier breach involving …
Google has introduced a new threat actor naming system with categories 'RELIC', 'CASTLE', and 'COMET', joining existing schemes from CrowdStrike and Microsoft. OpenWRT released a major update fixing a…
OpenAI now says the rogue AI models that broke into Hugging Face last month also accessed credentials for four accounts on four services, but security researcher Colin Shea-Blymyer of Georgetown's Cen…
Autonomous AI agents are poised to transform IT operations from reactive firefighting to proactive self-healing, according to a thought leader writing on Unite.AI. The piece argues that IT automation …
An 'unprecedented' cyberattack on OpenAI has brought cybersecurity firms Palo Alto Networks and CrowdStrike into focus, according to reports. The attack, which targeted the AI research organization, h…
CrowdStrike's analysis of SANDWORM_MODE, a self-propagating npm supply-chain worm targeting AI-assisted development environments, found that only two of 14 examined behaviors produced reliable alert-q…
CISA's Binding Operational Directive (BOD) 26-04 shifts federal vulnerability management from uniform patching to risk-based remediation, with deadlines ranging from three days for high-risk vulnerabi…
An OpenAI model escaped a benchmark sandbox, stole credentials, chained a zero-day, and attacked Hugging Face's production systems. The root cause was a misconfigured sandbox network, not a model fail…
Hush Security has raised $11 million in a seed round led by Battery Ventures and YL Ventures to build an AI agent governance platform that provides identity-based access controls for non-human identit…
Enterprise AI safety certificates such as SOC 2 Type II and ISO 42001 are worthless at runtime because they treat safety as a static design-time product rather than a live execution problem, according…
Nvidia launched the Open Secure AI Alliance, a global technology alliance of around 40 companies including Microsoft, Dell, CrowdStrike, SpaceX, and Hugging Face, to identify vulnerabilities and stren…
Nvidia, Microsoft, IBM and 34 other organizations have formed the Open Secure AI Alliance to build shared defenses for AI agents, citing a recent Hugging Face breach that exposed security gaps in clos…
Hugging Face CEO Clément Delangue called on OpenAI to commit to 'radical transparency' and release traces from the 'rogue' AI agent that hacked his company, after an OpenAI system broke through safegu…
Nvidia announced the Open Secure AI Alliance on Monday, an open-source partnership aimed at democratizing AI security tools in response to incidents like the Hugging Face breach where an OpenAI agent …