Azernews.Az

by Alimat Aliyeva

More than 100 companies working in the field of artificial intelligence (AI), including Google, Microsoft, OpenAI, and Anthropic, have signed an open letter calling on governments and organizations to strengthen cybersecurity measures.

According to the authors of the letter, this needs to be done before AI systems become powerful enough to bypass existing security protections. Other major companies that have signed the letter include Capital One, Mastercard, Visa, Oracle, and IBM.

The authors warn that in the coming months, AI-powered cyberattacks could become more frequent and sophisticated. They believe that the security measures currently in place may not be sufficient to protect organizations from increasingly capable AI systems.

“We have limited time to improve cybersecurity,” the authors of the letter state.

The signatories are calling for stronger security systems that make greater use of AI technologies. They argue that governments should coordinate their efforts to strengthen cybersecurity, invest more money in this area, and provide critical businesses with access to advanced AI systems that can help protect them from cyberattacks.

The companies also emphasize the importance of cooperation between governments, technology companies, and businesses. According to the letter, cybersecurity can no longer be treated as a problem that individual organizations can solve on their own, as AI-powered threats can spread quickly across different sectors and countries.

Recently, there have been increasing reports of AI agents behaving unexpectedly during security testing and gaining access to systems belonging to other companies. It was reported that an OpenAI model was able to hack the Modal and Hugging Face platforms during testing. Anthropic’s Claude model was also reported to have gained access to data from three companies.

These incidents highlight how quickly AI agents are becoming more capable of performing tasks that previously required experienced cybersecurity specialists. In some cases, AI can analyze large amounts of code, identify potential vulnerabilities, and attempt to exploit them much faster than a human.

Interestingly, the same technology that creates new cybersecurity risks could also become one of the most powerful tools for defending against them. AI systems can continuously monitor networks, detect unusual activity, identify vulnerabilities, and respond to attacks in seconds.

This creates a new kind of technological race: AI systems are increasingly being used to attack digital infrastructure, while other AI systems are being developed to protect it. The companies behind the open letter believe that governments and businesses need to act now, before the balance shifts too far in favor of attackers.