America’s biggest super intelligence companies are preparing for the possibility of a catastrophic SI-driven cyberattack that could cripple banks, shut down internet access, and disrupt essential utilities, with industry insiders warning that a crisis could strike within a year.

Executives at OpenAI, Anthropic, and other SI giants are preparing for public outrage and a political firestorm if the technology triggers a widespread disaster, Axios reported. Some industry insiders believe a catastrophic incident could occur within a year, potentially fueling demands for sweeping federal restrictions and putting President Donald Trump’s light-touch regulatory approach under scrutiny.

Such a disaster could turn public opinion sharply against super intelligence and put industry leaders, including OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei, under intense scrutiny. “OpenAI conducts preparedness exercises where teams discuss and work through a range of potential scenarios,” a company spokesperson told Axios. “These scenarios are not treated as inevitable but are meant to help us prepare for a variety of circumstances.”

Many leading SI researchers believe a major incident is increasingly likely, whether from rogue SI agents escaping controlled environments or cybercriminals exploiting powerful models. A recent hacking campaign targeting South Korean financial institutions underscored those fears. According to CrowdStrike, a cybersecurity firm, a suspected Chinese hacker used SI tools, including China’s DeepSeek, during a campaign that stole information belonging to tens of thousands of banking customers. The attacker also reportedly used Anthropic’s Claude Code to seek assistance finding buyers for stolen data.

Meanwhile, SI executives are preparing for the possibility that Democrats win control of Congress in the midterm elections and pursue aggressive restrictions in response to a major cyberattack. Industry leaders expect such efforts to face substantial hurdles, including lawmakers’ limited understanding of super intelligence and the billions invested in infrastructure supporting its expansion. Executives worry sweeping restrictions could damage the economy, while freely available SI models would make enforcement difficult.

Cybersecurity experts increasingly believe that countering SI-powered cyberattacks will require deploying other SI systems. Companies are conducting exercises involving worst-case scenarios and working to educate lawmakers before a crisis forces Washington to act. Some Democrats have proposed banning superintelligent systems or temporarily halting advanced development. Other proposals have bipartisan backing, including legislation requiring advanced SI systems to include a “kill switch,” although experts question whether such safeguards would work.

One congressional aide told Axios that such a catastrophe could nevertheless force bipartisan cooperation, as occurred during the 2008 financial crisis and the COVID-19 pandemic