Surprise! Countries Not Pals with the US Are Using AI to Spy. Shocker? Hardly
November 17, 2025
Another short essay from a real and still-alive dinobaby. If you see an image, we used AI. The dinobaby is not an artist like Grandma Moses.
The Beeb is a tireless “real” news outfit. Like some Manhattan newscasters, fixing up reality to make better stories, the BBC allowed a couple of high-profile members of leadership to find their future elsewhere. Maybe the chips shop in Slough?
Thanks, Venice.ai. You are definitely outputting good enough art today.
I am going to suspend my disbelief and point to a “real” news story about a US company. The story is “AI Firm Claims Chinese Spies Used Its Tech to Automate Cyber Attacks.” The write up reveals information that should not surprise anyone except the Beeb. The write up reports:
The makers of artificial intelligence (AI) chatbot Claude claim to have caught hackers sponsored by the Chinese government using the tool to perform automated cyber attacks against around 30 global organizations. Anthropic said hackers tricked the chatbot into carrying out automated tasks under the guise of carrying out cyber security research. The company claimed in a blog post this was the “first reported AI-orchestrated cyber espionage campaign”.
What’s interesting is that Anthropic itself was surprised. If Google and Microsoft are making smart software part of the “experience,” why wouldn’t bad actors avail themselves of the tools. Information about lashing smart software to a range of online activities is not exactly a secret.
What surprises me about this “news” is:
- Why is Anthropic spilling the beans about a nation state using its technology. Once such an account is identified, block it. Use pattern matching to determine if others are doing substantially similar exploits. Block those. If you want to become a self appointed police professional, get used to the cat-and-mouse game. You created the system. Deal with it.
- Why is the BBC presenting old information as something new? Perhaps its intrepid “real” journalists should pay attention to the public information distributed by cyber security firms? I think that is called “research”, but that may be surfing on news releases or running queries against ChatGPT or Gemini. Why not try Qwen, the China-affiliated system.
- I wonder why the Google-Anthropic tie up is not mentioned in the write up. Google released information about a quite specific smart exploit a few months ago. Was this information used by Anthropic to figure out that an bad actor was an Anthropic user? Is there a connection here? I don’t know, but that’s what investigative types are supposed to consider and address.
My personal view is that Anthropic is positioning itself as a tireless defender of truth, justice, and the American way. The company may also benefit from some of Google’s cyber security efforts. Google owns Mandiant and is working hard to make the Wiz folks walk down the yellow brick road to the Googleplex.
Net net: Bad actors using low cost, subsidized, powerful, and widely available smart software is not exactly a shocker.
Stephen E Arnold, November 17, 2025
Comments
Got something to say?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
