CrowdStrike
4 mentions across all digests
Cybersecurity company listed among the 40+ partners deploying Anthropic's Mythos model for defensive security work and vulnerability scanning under Project Glasswing.
Next.js developer Vercel warns of customer credential compromise
Context.ai's March OAuth token theft gave attackers a backdoor into Vercel's Google Workspace, exposing customer credentials due to overly broad permission grants.
Exclusive: Artemis raises $70M to help fight AI-powered attacks with AI
AI security startup Artemis raises $70M Series A to defend against AI-powered cyberattacks, with early clients including Mercury, Wix, and Lemonade already on track for multi-million ARR.
The Closing of the Frontier
Anthropic restricts its Mythos frontier model to enterprise-only access, marking a shift from open AI exploration to commercial gatekeeping that risks widening advantage for well-funded players.
Anthropic debuts preview of powerful new AI model Mythos in new cybersecurity initiative
Anthropic deploys Mythos, a frontier AI model, across 40+ tech partners including Amazon and Apple for cybersecurity, discovering thousands of zero-day vulnerabilities in weeks.
A major enterprise security vendor (CrowdStrike, Palo Alto Networks, or Fortinet) will announce a 'read-only AI' or 'least-privilege AI agent' product tier within 8 weeks, explicitly restricting AI security tools to observation-only mode by default, with write access requiring human-in-the-loop approval.
At least one major enterprise security vendor (CrowdStrike, Palo Alto Networks, or SentinelOne) will announce an AI-powered security tool integrity verification product — specifically designed to detect when defensive/trusted software has been compromised or weaponized — within 6 weeks.