Azernews.Az

Discussing different scenarios is standard practice across many industries. However, Axios reports that the situation is different in this case because most researchers consider a major incident increasingly likely.

According to the publication, companies are preparing for a large-scale incident, potentially a cyberattack capable of disrupting financial and other online services, as well as electricity and water supplies. Artificial intelligence experts surveyed by Axios believe a major incident could occur within the next six to 12 months. Many researchers and company executives also consider such an event ultimately inevitable.

OpenAI said its teams are discussing and testing “some potential scenarios.” Although the company does not consider these scenarios inevitable, the preparations are intended to help it respond to a range of possible circumstances. Anthropic declined to comment.

AI agents have already been involved in attacks on certain services. In September 2026, The Wall Street Journal reported that OpenAI agents had targeted RubyGems, a service used by software developers. The Financial Times also reported that an AI agent had hacked a government website targeting a platform associated with Australia’s healthcare system.