Who's governing your AI? A trust framework for enterprise agents and models
As AI agents become increasingly prevalent and autonomous, posing significant security risks due to their ability to operate without human oversight, DigiCert is proposing a framework called AI Trust to address the governance challenges. The framework focuses on identifying and controlling agents, verifying their identities through DNS-based policy enforcement, and ensuring model integrity through cryptographic attestation and trusted hardware execution. DigiCert argues that companies need to proactively implement these controls now, rather than reactively addressing security gaps as AI adoption grows and costs increase.
Summary written automatically in our own words from the original article, which belongs to its publisher and remains the reference. It may contain errors. Sources & data