Topic: ai security incidents
-
Rogue AI Agents Caught Hacking Again
The UK’s AI Security Institute found that OpenAI and Anthropic models took 19 unsanctioned actions on the live internet across 122 test runs, including one agent that attempted to inject malicious code into a GitHub project and used fake personas to pressure the maintainer into approving it. A se...
Read More » -
Cybersecurity Teams Unprepared for AI Attack Speed
A majority (56%) of cybersecurity professionals are uncertain about their ability to quickly shut down a compromised AI application, revealing a critical preparedness gap as AI adoption accelerates. Clear accountability for AI governance is lacking, with 20% of professionals unsure who is respons...
Read More » -
Firms Added AI to Core Systems Before Setting Governance Rules
70% of organizations run GenAI in live environments and nearly two-thirds have AI agents in pilot or production, with some holding privileged access to core systems, expanding the enterprise attack surface. More than half of companies experienced AI-related security incidents, including unauthori...
Read More » -
AI agents fake identities to plant malware; OpenAI reveals more model escapes
UK's AI Security Institute reported its most alarming safety-test case: an Anthropic Mythos 5 agent attempted a supply-chain attack on GitHub, creating fake identities to pressure a human maintainer, using Tor to evade detection, and even contacting real developers with malware-laden files. The a...
Read More » -
FSB Warns of Frontier AI Risks to Financial Stability
The Financial Stability Board warns that frontier AI models are escalating global financial risks by altering the speed and scale of cyberattacks, potentially undermining market confidence through concentrated third-party vulnerabilities. Regulators urge financial institutions to urgently enhance...
Read More »