CrowdStrike's AI Security Coalition
- CrowdStrike said on April 23 it launched Project QuiltWorks, a coalition with OpenAI, Accenture, EY, IBM Cybersecurity Services and Kroll to find and fix software flaws frontier artificial intelligence models are surfacing in production code. - CrowdStrike said the coalition pairs OpenAI and Anthropic models with Falcon threat intelligence and a partner network of more than 10,000 certified professionals for assessments, prioritization, board reporting and guided remediation. - The push follows warnings that AI can sharply speed vulnerability discovery, shrinking patch windows from monthly cycles toward continuous response. (markets.ft.com)
CrowdStrike has launched Project QuiltWorks, a coalition with OpenAI, Accenture, EY, IBM Cybersecurity Services and Kroll to tackle software vulnerabilities uncovered by frontier artificial intelligence models. (markets.ft.com) The company announced the effort on April 23, 2026, saying the group will assess, prioritize and continuously remediate flaws in production code that newer AI systems are finding faster than older tools did. (markets.ft.com) CrowdStrike also introduced a Frontier AI Readiness and Resilience Service, which it said will deliver ongoing scans, expert prioritization and guided remediation directly to customers. (markets.ft.com) (crowdstrike.com) The underlying problem is simple: stronger AI systems can read large amounts of code and spot logic bugs, design flaws, misconfigurations and unusual exploit paths at a pace human reviewers and conventional scanners struggle to match. (markets.ft.com) CrowdStrike said its Falcon platform, which it says processes trillions of security events daily, will be used to rank which newly found flaws are actually reachable and exploitable by attackers, rather than just theoretically dangerous. (markets.ft.com) The company said QuiltWorks extends that triage and repair work through a partner network of more than 10,000 certified professionals, with board-level risk reporting and remediation help inside customer environments. (markets.ft.com) CrowdStrike executive Daniel Bernard told CRN that the launch followed Anthropic’s disclosure that advanced AI tools could dramatically increase vulnerability discovery, setting off demand from customers for guidance. (crn.com) Bernard said one participating company had already found 45 million vulnerabilities with the new capabilities, and he predicted patching would shift from “Patch Tuesday” toward constant repair over the next six to 12 months. (crn.com) He also said the project is meant to be less dependent on any single model, arguing that the way companies operationalize multiple models and remediation workflows now matters more than choosing one artificial intelligence provider. (crn.com) For enterprises, the pitch is not just better detection. It is a standing process for deciding which AI-discovered flaws to fix first, who fixes them, and how to show boards that exposure is being reduced. (markets.ft.com) The closing argument from CrowdStrike is that the window between discovery and exploitation is narrowing, so the security industry is trying to turn AI-driven bug hunting into faster, coordinated patching before attackers do the same. (markets.ft.com)
