Gates Warns That AI Cyberattacks Will Bring "a Lot of Negative Surprises" in the Next Year

Bill Gates said Sunday that the most immediate danger from artificial intelligence is not a system escaping human control but criminals using the latest models without safeguards, and that governments are not doing enough to prepare.

"When you have bad people, no safeguards, and then the latest models, that's where, over the next year, primarily in the cyber area, we are going to have a lot of negative surprises," Gates told Jake Tapper on CNN's State of the Union. He said the front that "we'll probably see most dramatically in the next year is cyberattacks, because there, there's a known business model where, you know, you threaten to do something and ask somebody to pay for that, or you create a disruption that changes something in the stock market or some betting thing that you go and do."

Asked whether government leaders were doing enough, Gates said, "Well, no, we're not doing enough." He separated the near-term risk from the longer one. "The real danger right now is not the AI is getting out of control. We need to work on that, but that's years away in terms of the big impact that might have."

The warning sits next to a September essay by Anthropic chief executive Dario Amodei that called for the industry to slow capability gains. Elon Musk replied on X, "Dario is right." OpenAI's Sam Altman wrote, "I agree with Dario that we need to pace the frontier," and said OpenAI would give independent evaluators "employee-like access." Gates's CNN comments address a different horizon: not a pause on frontier training, but the use of models that already exist by people who remove the guardrails.

CrowdStrike's 2026 Global Threat Report, cited in coverage of the interview, said attacks by AI-enabled adversaries rose 89 percent year over year in 2025, and that the average time for criminals to move laterally through a compromised network fell to 29 minutes. Those figures are the security firm's, not Gates's. They describe the environment he was pointing at.

Gates has spent the past month on two AI messages. The Gates Foundation pledged at least $1 billion over two years for AI tools in health, education and agriculture in lower-income countries, and he wrote that decisions in the next 12 to 18 months would determine whether the technology reaches "those who have the least." On Sunday he used the same 12-month window for a different outcome: extortion and market disruption by people who strip the safeguards off the newest models. He did not name a company, a vulnerability or a date. He named a business model that already exists and a set of tools that make it faster.