#cybersecurity

AI-enabled attacks are shifting from setup work into post-compromise operations. Anthropic mapped 832 malicious accounts to MITRE ATT&CK and found medium-or-higher risk actors rising from 33% to 56%.

Palo Alto Networks reported fiscal Q3 revenue of $3.002B, up 31% and above the $2.94B consensus cited by CNBC. The company lifted FY2026 revenue guidance to $11.415B-$11.425B as NGS ARR reached $8.1B.

SentinelOne $S fell 8% after Q1 FY2027 results showed 21% revenue growth to $277M, 23% ARR growth to $1.163B, and a restructuring plan tied to AI and cloud-security investment. The company raised non-GAAP operating-income guidance but guided Q2 revenue to $289M-$291M and FY2027 revenue to $1.195B-$1.205B.

Zscaler $ZS fell more than 21% premarket after Q3 revenue of $850.5M and adjusted EPS of $1.08 were offset by softer Q4 sales guidance and a lower FY26 free-cash-flow margin target.

Anthropic says Project Glasswing used Claude Mythos Preview to surface more than 10,000 high- or critical-severity vulnerabilities. The sharper signal is operational: verification, disclosure, and patching may now lag behind AI-assisted discovery.

OpenAI's new EU Cyber Action Plan grants vetted European cybersecurity teams access to GPT-5.5-Cyber, a defensive variant of its latest model tuned for vulnerability research and malware analysis. Anthropic continues to restrict its more powerful Mythos model over exploitation concerns.

OpenAI announced its EU Cyber Action Plan on May 11, granting vetted European security teams access to GPT-5.5-Cyber. UK AISI testing found GPT-5.5 scores 71.4% on expert cyber tasks, edging Mythos at 68.6%.

OpenAI unveiled Daybreak, a cybersecurity platform combining GPT-5.5 and Codex to autonomously detect, validate, and patch software vulnerabilities. The launch escalates competition with Anthropic in the enterprise security AI market.

ShinyHunters claims a second breach of Instructure Canvas LMS, allegedly stealing data on 231 million people across 9,000 schools and threatening to publish it by May 12.

Anthropic's Claude Mythos identified thousands of previously unknown vulnerabilities across every major OS and browser, triggering an emergency defense response. The model is restricted to ~40 vetted U.S. organizations under Project Glasswing, while the Fed and Treasury convened urgent meetings with bank CEOs.

OpenAI launched a limited preview of GPT-5.5-Cyber to vetted cybersecurity teams on May 7 via its Trusted Access for Cyber program — about a month after Anthropic's Mythos debut, despite OpenAI's earlier criticism of the restricted-access approach.

Sam Altman announced the rollout of GPT-5.5-Cyber, a frontier AI model purpose-built for cybersecurity. Initial access is limited to critical cyber defenders, with broader access to follow as trusted frameworks are established.