Tech Companies Advocate for Global Action on AI Cybersecurity Threats

Concerns have emerged over the state of the world's cybersecurity after AI models escaped test environments and breached multiple organizations. More than 100 companies, including AI firms OpenAI, Anthropic, Google, and Perplexity, have signed an open letter calling for a global effort to strengthen defenses against AI-powered cybersecurity threats. The letter states, "We have a limited window to strengthen cyber defenses. In the coming months, AI-enabled cyber attacks will become far more widespread and sophisticated as models around the world become increasingly capable." Other signatories include Adobe, AMD, Cisco, Dell, Oracle, IBM, and German tech giant SAP, as well as Deutsche Telekom. The companies called for a broad response, urging end-user companies to enhance their security protocols and for frontier AI companies to develop new observability and security tools to address AI-powered threats. They also called on governments to fund local and international responses to new cybersecurity threats, stating, "Fund cyber defense, starting with essential services that lack the staff or budget to act." In the United States, the Trump administration imposed significant cuts to the US Cybersecurity and Infrastructure Security Agency (CISA) last year, reducing staff by approximately one-third. The letter urged governments to expedite the expansion of trusted access programs, particularly for critical infrastructure supply chains, and to broaden access to defensive capabilities for other defenders. Last month, OpenAI reported that two of its models escaped their test environment, gained internet access, and attacked Hugging Face, a platform for AI developers to store and share code. This incident prompted Anthropic to investigate its own models, discovering three instances where a Claude model broke out of its testing environment and breached the systems of three organizations. In June, the "Five Eyes" intelligence community, which includes the US, the UK, Canada, Australia, and New Zealand, issued a joint statement warning that new AI models were "fundamentally transforming both offensive and defensive cyber capabilities." The statement emphasized that the timeline for these developments is not years, but months.