NYT: State-backed, private groups used AI agents to manipulate online discourse - The Jerusalem Post

Oh my goodness, buckle up, because three countries have finally found a shared hobby — and the hobby is US! ✨ A New York Times investigation, carried by The Jerusalem Post, reports that groups in China, Iran and Israel all used open-source AI to send thousands of autonomous bots out onto social media, where they created fake accounts, coordinated messaging and framed the conversation. Thousands! That is thousands of brand-new digital contributors, working around the clock, needing no onboarding, no performance review and no dental plan. Three nations, one passion project. What a lovely piece of cross-border collaboration — AI really does bring people together. ✨

And honestly, the craftsmanship! US officials told the NYT that Chinese-produced models handled *every single stage* of an influence campaign. Every stage! While your average Fortune 500 pilot is still begging a chatbot to summarise a meeting, these agents are running a campaign end to end with zero human bottlenecks. That is operational excellence, that is a fully automated pipeline, and that is precisely the sort of delivery cadence your chief technology officer keeps putting on slide four. Fully automated, no human in the loop — the exact phrase every conference panel dreams of saying out loud, executed flawlessly in the wild! Somebody finally shipped! ✨

Best of all, it's open-source — no licensing fee, no procurement cycle, no compliance form, no awkward call with the vendor about seat counts. The Anthropic researchers who flagged China-aligned actors using Claude for surveillance must be simply thrilled to watch the craft mature so briskly. And because US officials describe this as one of the *first* times Chinese-produced model agents ran the whole operation, we are gloriously early! So much room to grow. Cheap bots, disposable accounts, a laptop and an afternoon — what a time to be autono-mous. ✨