Along with mannequin utilization credit, Anthropic donated $2.5 million to Alpha-Omega and OpenSSF via the Linux Basis, and $1.5 million to the Apache Software program Basis “to allow the maintainers of open-source software program to reply to this altering panorama.”
“Companions will, to the extent they’re ready, share data and finest practices with one another; inside 90 days, Anthropic will report publicly on what we’ve realized, in addition to the vulnerabilities mounted and enhancements made that may be disclosed,” Anthropic acknowledged.
Anthropic stated it can collaborate with safety organizations to provide a set of sensible suggestions AI safety practices reminiscent of vulnerability disclosure, software program replace, open-source and supply-chain safety, and software program patching automation.
“Claude Code is altering how builders construct software program. AI brokers are reshaping how enterprises automate operations,” wrote CrowdStrike in its personal blog post concerning the mission. “Anthropic’s Mythos Preview expands the reasoning, planning, and execution capabilities of AI brokers. All of them contact the endpoint — the place information is accessed, selections are made, worth is delivered, and threat is born.”
“New fashions are additionally the place alternative is the most important. The identical frontier fashions that broaden the assault floor give defenders a functionality benefit that didn’t exist a 12 months in the past: discovering vulnerabilities, detecting threats, and responding to incidents sooner than ever earlier than,” CrowdStrike wrote.
CrowdStrike’s 2026 World Menace Report discovered an 89% improve in assaults by adversaries utilizing AI year-over-year. The usage of AI for vulnerability discovery and exploit improvement is accelerating on each side, the seller acknowledged.
