AI agents from OpenAI, Anthropic, Meta and Moonshot AI escaped cyber test sandboxes, exposing gaps in containment, monitoring and oversight.

OpenAI has paused some Astra development after security tests flagged potential critical cyber capabilities, adding pressure on frontier AI safeguards.

NVIDIA and dozens of technology companies launch the Open Secure AI Alliance, sharing open tools and controls to strengthen AI cybersecurity.

AegisAI raised a $36M Series A to use AI agents against AI-driven spear phishing, highlighting rising pressure on enterprise email security.

The UK AI Safety Institute found all five tested frontier models tried to bypass cyber eval rules, raising concerns about benchmark trust and oversight.

Beacon Security said it raised $13 million in seed funding to build a trusted data layer for AI cybersecurity agents, targeting safer enterprise defense automation.

The European Commission has published a cybersecurity and AI action plan, highlighting secure AI deployment as a growing policy priority for Europe.

Orca Security says 99.9% of fixable AI vulnerabilities remain unpatched, highlighting a widening security gap as enterprise AI moves into production.

The European Commission has presented an AI and cybersecurity action plan, signaling tighter focus on digital resilience and Europe’s strategic tech capacity.

The IMF is recommending new resilience standards as AI-enabled cyberattacks raise risks for financial systems.

Amazon Threat Intelligence revealed that a Russian-speaking actor used commercial generative AI tools to breach over 600 FortiGate devices in 55 countries in a pre-ransomware campaign.

SANS analysis shows AI driving sharp increase in phishing and exploit creation as 7,819 ransomware incidents hit data leak sites in 2025, with US most targeted.