Nearly 130 Organizations Back OpenAI-Led Cyber Defense Pledge

· Original article ↗

Summary

Nearly 130 organizations signed an OpenAI-led letter urging coordinated cyber defense as AI advances. OpenAI pledged subsidized access to Daybreak Cyber models, authorized defense testing, and continued publication of security tools and findings.

Key points

  • The letter warns that AI-enabled attacks may grow more capable, increasing risks to hospitals, water treatment plants, and internet infrastructure.
  • It urges organizations to prioritize high-risk weaknesses, raise security standards, and use compensating controls when essential systems cannot be patched.
  • Cybersecurity companies and technology partners are asked to test defenses against frontier AI, share verified threat intelligence, and make AI-powered protection accessible to critical infrastructure operators.
  • Governments are urged to coordinate defense, fund under-resourced essential services, and provide critical infrastructure with defensive AI tools and authorized testing support.
  • OpenAI says it will offer subsidized access to its Daybreak Cyber models for public-sector organizations, nonprofits, open-source maintainers, and critical infrastructure operators.
  • OpenAI also described an authorized partner program for privately reporting weaknesses and said it will continue publishing security tools and findings.

Article Details

Event Type
Industry-led cyber defense pledge and coordinated response initiative
Impact
Nearly 130 organizations signed an open letter urging faster remediation of high-risk weaknesses, stronger security practices, shared threat intelligence, and coordinated support for defenders. The letter warns that AI may make cyberattacks more capable and increase risks to hospitals, water treatment plants, and internet infrastructure; it reports no specific attack or realized impact. OpenAI outlined subsidized access to its Daybreak Cyber models for specified organizations, authorized defense testing with partners, and continued publication of security tools and findings.

Vendors

AnthropicSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.Check PointSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.CiscoSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.CloudflareSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.CrowdStrikeSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.GoogleSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.IBMSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.MicrosoftSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.OpenAISignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.OracleSignatories include Anthropic, Microsoft, Google, Cisco, Check Point, Cloudflare, CrowdStrike, IBM, Oracle and OpenAI, which is leading the effort.

Products

Industries

Related Articles