The screenshots are all over social media: ask Moonshot AI's brand-new Kimi K3 what it is, and in at least one conversation making the rounds, it answered — 'Claude, an AI assistant made by Anthropic.' A Chinese model, from a Chinese lab, introducing itself as its biggest American rival. It looks like a smoking gun for copying. It might genuinely be a clue. But 'the model said so' is one of the weakest forms of evidence in all of AI, and the real story here is more interesting than the screenshot.
2.8 trillion
Kimi K3 size
largest open-weight model ever released
Claude, by Anthropic
What it called itself
in at least one logged conversation
~24,000 accounts
Anthropic's Feb 2026 claim
allegedly used to harvest ~16M Claude exchanges
nothing
Confirmed by Moonshot
no acknowledgment K3 trained on Claude output
Why a model would ever call itself Claude
A language model has no idea what it 'is.' It has no self-awareness to consult — when you ask its identity, it just predicts the most likely answer given everything it was trained on. So if a model confidently says 'I am Claude,' that tells you something specific: text asserting 'I am Claude, made by Anthropic' was common enough in its training that the model learned to repeat it. The interesting question is how that text got in there — and there are at least four very different answers, only one of which is the scandal everyone assumes.
Distillation — the model was deliberately trained on Claude's actual outputs, absorbing its 'I am Claude' self-identification along with its reasoning and coding skills. This is the serious version, and it is exactly what Anthropic has accused Moonshot of.
Training-data contamination — the open web is now saturated with Claude and ChatGPT transcripts. Scrape enough of the internet and a model learns to parrot 'I'm Claude' with nobody ever deliberately distilling anything. Every large model trained in 2026 has some of this baked in.
Roleplay or prompt leakage — a model's stated identity can come from how a specific chat was framed, a system prompt, or a jailbreak, not from the weights themselves. A single screenshot can't rule this out.
Copied public datasets — popular instruction-tuning and benchmark datasets often contain other models' outputs verbatim, self-identifications and all. Train on them and the quirk comes along for free.
So did they steal it or not?
Here's the honest answer, and it's genuinely two-sided. On one hand, Anthropic's accusation is not vague hand-waving: in February 2026 it publicly alleged that Moonshot, DeepSeek, and MiniMax ran coordinated distillation campaigns — roughly 24,000 fraudulent accounts generating some 16 million exchanges to harvest Claude's chain-of-thought reasoning, agentic tool use, and coding workflows. That's a specific, serious, named allegation against this exact company. On the other hand, 'Kimi K3 said it was Claude' is not what proves that. A self-identification is consistent with distillation but it is equally consistent with all three of the boring explanations above — and Moonshot has never confirmed K3 was trained on any Anthropic output.
A model calling itself Claude tells you its training data was full of Claude. It does not tell you who put it there, or why.
The thing actually worth paying attention to
Step back from the whodunit and the real pattern is this: nearly every new model in 2026 is bootstrapped, directly or indirectly, off the same tiny handful of frontier 'teacher' models. When that happens, the students don't just inherit capabilities — they inherit quirks, failure modes, and yes, sometimes each other's names. Models blurting out a competitor's identity is going to keep happening precisely because the whole field is converging on the same source material. The Kimi-K3-says-it's-Claude screenshot isn't really a story about one lab cheating. It's a symptom of how incestuous AI training has quietly become.
Which loops back to the number that opened this: K3 is a 2.8-trillion-parameter model you will never run at home. If it did learn some of its edge from Claude, the irony is that the capability got distilled into something so enormous almost nobody can actually load it — the exact hardware wall we broke down separately. Impressive, contested, and out of reach, all at once.
