AI Leaders Warn of a Closing Window to Stop Dangerous Cyberattacks

Some of the biggest names in technology are sounding an alarm about a serious threat. Leaders from OpenAI, Anthropic, Google, and Microsoft signed an open letter warning that artificial intelligence could soon be used to carry out dangerous cyberattacks. These attacks could target hospitals, banks, and other important organizations. The letter said there may be only a few months left to build stronger defenses before that threat becomes a reality.

The letter was published on a Thursday and signed by leaders from dozens of companies. Along with the big tech firms, security companies like CrowdStrike and banks such as Citi and Capital One also added their names. All of them agreed that action needs to happen soon. They described the current moment as a 'limited window' — a short period of time when defenders can still get ahead of the problem.

A recent report from CrowdStrike found that AI-powered attacks jumped by 89% in 2025 compared to 2024. That means attackers are already starting to use AI as a weapon. The letter warned that the old ways of handling cybersecurity 'won't be enough' going forward. New and stronger tools are needed to keep up with the growing danger.

The letter pointed out several weak spots that leave computer systems open to attack. These include old software that has never been fixed, weak passwords, and systems that have not been updated in a long time. Security teams need more money and better tools to fix these problems. The letter said defenders must be equipped with the most advanced AI technology available.

One important idea in the letter was that companies and experts should share what they know. If one organization figures out how to stop a certain type of attack, they should tell others. The letter said groups should 'share threat intelligence and tested playbooks.' This means sharing plans and information so that more organizations can stay safe, not just the ones with the most resources.

The letter also called on governments to play a bigger role. Leaders said local, national, and international governments should work together and fund cybersecurity programs. Every organization, big or small, should make protecting its computer systems a top priority. That might mean replacing old equipment or upgrading software that is no longer safe to use.

Finally, the letter said that leading AI companies have a special responsibility. Since they build the most powerful AI tools, they must also help keep those tools from being misused. That includes training more cybersecurity workers, giving responsible access to their AI models, and making sure the technology itself is kept secure. The message from all the signatories was clear: the time to act is now.