Gates: AI-Driven Cyberattacks Pose Imminent Threat

Bill Gates, co-founder of Microsoft, told Agence France-Presse in a recent interview that AI-driven cyberattacks are likely to increase over the next year. "The front we will probably see most visibly increase next year is cyberattacks," he said, placing immediate emphasis on malicious use rather than runaway systems.

Gates warned that the more distant but still important risk of AI escaping human control remains a concern, yet its broad consequences are unlikely to be seen for several years. "The real danger at present is not that AI gets out of control. We should work on that, but the big impacts are still a few years away," he said.

His urgent worry is different: bad actors exploiting advanced models and removing safety limits. "Right now, you are allowed to remove all these safety measures," Gates said, highlighting how accessible modifications can make systems harmful.

He cautioned that the combination of malicious intent, weak or absent safeguards and access to the latest AI models could produce unforeseen negative consequences next year, particularly in the realm of cyberattacks.

Limits of Industry Self-Regulation

Gates also rejected the notion that AI companies can reliably police themselves. He said that this view had also been voiced by Donald Trump, the president of the United States. Gates argued that relying on voluntary oversight by companies is inadequate for a technology with such far-reaching potential.

"The idea that for the most dangerous technology in human history the companies should have the final say and that the companies themselves are the only regulatory factor for society shows that our actions are not sufficient."