More than a hundred leading technology sector companies, including OpenAI, Anthropic, Microsoft, Google, and Amazon Web Services, have signed an open letter warning of an impending wave of cyberattacks using artificial intelligence. In my assessment, this is not just another declaration—it is a signal that the digital threat landscape is changing fundamentally and faster than many expect.
The letter was signed by representatives not only from IT giants but also from the banking sector, insurance companies, and semiconductor manufacturers. Key signatories include CrowdStrike, Cloudflare, and Palo Alto Networks. Such broad coverage indicates that the threat is being taken seriously at all levels of the economy.
The essence of the letter: four fronts of defense
The authors of the document identify four critical groups that need to act immediately. First, any company must elevate cyber defense to the level of senior management and, above all, close the most dangerous vulnerabilities. Second, cybersecurity solution providers must make AI protection accessible to operators of critical infrastructure.
Special attention is given to government agencies—they are urged to coordinate response measures and help protect hospitals, utilities, and local authorities. Leading AI companies are recommended to open access to models and fund training for those who lack resources for defense.
"Modern advances in AI are already giving defenders new ways to eliminate weaknesses that have accumulated over the years. If we act decisively, we can use this window of opportunity to make the digital world much safer," the letter emphasizes.
Data confirms: the threat is already here
Signs of the impending threat had been recorded earlier. Taiwan-based TeamT5, which specializes in tracking cyber espionage, found that China-linked hacker groups doubled the number of attacks after switching to DeepSeek and other open models.
The hackers' choice is explained not by power but by low cost. In June, Anthropic provided a broader assessment: they analyzed 832 blocked accounts and found that the share of malicious actors with medium and high risk grew from 33% to 56% over the year. Notably, AI now independently handles privilege escalation and network movement—previously, technical skills were indispensable here. The level of an attack increasingly depends less on the attacker's expertise.
South Korea's Genians discovered that the North Korean intelligence group Kimsuky is actively testing local AI tools. TRM Labs rated the use of AI for criminal purposes at 54 out of 100 for this year—compared to 28 in 2024. In the first half of 2026, they counted 201 hacks of crypto projects, whereas only 83 were recorded a year earlier.
The letter notes that the current leap in AI development "could translate into more robust protection for everyone."
My comment: The rise in attacks on crypto projects from 83 to 201 in a year is not just statistics. It is direct evidence that malicious actors are actively arming themselves with AI, and the industry needs to respond symmetrically. The collective defense initiative is a step in the right direction, but the key question is the speed of implementation. The window of opportunity the letter's authors speak of may close faster than we think.