IC3 Researchers: Crypto's Role in Solving AI Trust and Payment Issues Is Limited

According to The Block, on June 9, IC3 researchers from Cornell University, Carnegie Mellon University, and other leading institutions released a report stating that claims about crypto solving AI's biggest challenges are often overstated, with crypto's effectiveness in addressing AI trust and payment issues being limited.

The report noted that providing AI agents with crypto wallets does not make them smarter or more resistant to human manipulation; wallets only enable automated transactions and on-chain infrastructure access without manual approval. While blockchain suits timestamping and registering specific digital content, it cannot determine how content was created and requires external tools for verification. If external tools fail, errors recorded on-chain become permanent. Regarding algorithmic bias, the researchers concluded it is unlikely to be resolved through decentralized AI, as bias stems from training processes and is typically mitigated through improved training or inference techniques; decentralization cannot address bias at its root, and no evidence supports that transparency and broader participation reduce bias or improve model outputs.