What happened
In a post on his own X account on October 6, Team Mirai leader Takahiro Anno said highly capable attack AI models are spreading, and that data leak incidents are mounting, so action needs to be taken quickly.
Why it matters
Anno is signaling that the spread of such AI models is no longer a future concern, and that responses need to speed up.
What to watch
Anno says the Diet statement has already come true, so the focus is on what concrete responses follow from here.
WHO IT HITSThis lands on security teams at companies and government bodies who may face more capable AI-driven cyberattacks, and on policymakers deciding what to do about data leaks.
Summaries like this, in your inbox every morning.
Anno's comments come as he says a prediction he made in the Diet in April has effectively materialized. He had said then that there were only a few months at most before models with Mythos-level attack capabilities would spread, referring to Anthropic's "Claude Mythos" model. He judged the performance of "GLM-5.3," an open AI model announced by Chinese AI developer Z.ai in August, to be almost Mythos-level, and said his statement had in fact come true.
The timing links a prior public warning to a concrete model release. Anno's concern is that such capabilities are spreading and that data leaks are already piling up, which is why he frames the need for action as urgent.
The stakes hinge on whether this warning translates into concrete policy or corporate security measures. For now, it is a public claim from an opposition party leader that a specific open model has brought a previously hypothetical risk much closer.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
