OpenAI, Anthropic, Amazon Web Services, Microsoft and more than 100 other companies have cautioned that organisations have only a few months left to shore up their defences against AI-enabled cyberattacks.
The warning was issued through a joint open letter, in which the signatories argued that hospitals, water treatment plants and other critical infrastructure face a growing wave of hacking attempts as artificial intelligence makes advanced cyber capabilities cheaper and easier for attackers to access, Axios reported.
A shrinking window
The letter states that the industry has 'a limited window to strengthen cyber defenses'. The coalition is pushing for coordinated action, urging organisations to use the time still available to secure critical infrastructure and raise the cost and difficulty of AI-assisted intrusions.
Signatories span cloud providers, cybersecurity firms, AI developers, telecom operators, financial services companies and think tanks, Axios noted.
What different players are being asked to do
The letter lays out a division of responsibilities across the ecosystem:
- Every organisation is being asked to tighten internal security standards, fix its highest-risk weaknesses, and raise the bar for the systems it buys, builds and deploys - including AI-generated code.
- Cybersecurity and technology companies are being urged to rapidly develop and roll out tools that make AI-powered defence usable for operators of critical infrastructure, while also sharing threat intelligence with one another.
- Governments are being called on to strengthen mechanisms for sharing actionable threat intelligence, coordinate defence efforts across local, national and international levels, and fund cyber-defence initiatives.
- Frontier AI companies are being asked to give defenders access to their most capable response models during major cyber incidents, along with funding, training and hands-on support, particularly for critical infrastructure operators.
A wave of AI-linked attacks
The letter arrives amid a string of attacks on critical infrastructure, including one that targeted US water systems using an apparently AI-generated exploitation script.
Vulnerabilities in the systems that run machinery at facilities such as water plants and power stations are not new, but they historically required attackers to spend significant time understanding the intricacies of those systems. Experts have told Axios that AI models are now sharply cutting down the time and effort hackers need to invest in that preparatory work.
Despite the scale of the warning, the signatories stopped short of pledging specific commitments, deadlines or investments as part of the letter.
