Enterprise Tech with Fexingo · 2026-07-20 · 9 min
Episode 123 of Enterprise Tech with Fexingo dives into a fast-growing clause in Fortune 500 software contracts: AI training data provenance. Lucas and Luna explore why companies like a major healthcare insurer recently demanded that their CRM vendor document every dataset used to train its generative AI features - from licensed sources to synthetic data to web-scraped content. They break down the legal risks: copyright infringement suits, regulatory exposure under GDPR's Article 22, and the reputational cost of unknowingly using training data from biased or toxic sources. Lucas walks through how procurement teams are adding 'data lineage exhibits' and audit rights, and why vendors are pushing back on these terms. The episode also touches on the practical challenge of proving provenance across third-party sub-processors. A concise, concrete look at how the Fortune 500 is rewriting the rules on AI training data - one contract at a time.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.