What if you could identify the AI model behind a chatbot without having access to its internal code? That's what makes LLMmap interesting. Researchers from @OneRSAC and @GeorgeMasonU developed a technique that can identify which of 42 LLMs is likely powering an interface, with over 95% accuracy, using as few as 8 interactions. Think about what this means for AI transparency, model attribution, and intellectual property. But identifying a model is only one part of the bigger picture. As AI models become easier to access, protecting the work behind them becomes just as important. This is where AI fingerprinting comes in. Sentient is exploring how fingerprints embedded in model behavior can help verify a model's provenance and support ownership investigations, even when models have been modified. AI is becoming more accessible, but builders still need ways to protect what they create. Open AI should also mean better ways to establish provenance and protect builders' rights.. @SentientAGI