Don't want to trade it yourself?
Our desk runs DEX portfolios on profit share.
Sentient Labs, an AI research company under the SENT Foundation, launched the open benchmark ‘Fingerprinting Index’ to assess the security of technologies used to verify the provenance and ownership of AI models, the company said.
Fingerprinting refers to a technique that embeds a unique behavioral signature into an AI model to prove its link to the original model even after unauthorized copying or modification. Sentient said its evaluation of 10 existing fingerprinting techniques found that most were vulnerable to sophisticated attacks. The company also released the open-source library “mlprints” so researchers can test new verification techniques and attack methods.