Lenz – A fact-checking API for AI-generated content (lenz.io)

Lenz has launched a fact-checking API designed specifically for AI-generated content, addressing the critical need for accuracy in AI applications that risk generating false information. This tool incorporates eight distinct models applying a five-stage verification process, ensuring claims are substantiated by real sources. Lenz guarantees independent multi-vendor support and includes mechanisms for opposing viewpoints, with three reviewers assessing each claim, thereby enhancing the credibility of its outputs. The significance of Lenz lies in its potential to improve the reliability of AI products, which often struggle with the problem of “hallucination,” or generating incorrect information. The API features four key functions—/extract, /assess, /verify, and /ask—allowing users to pull verifiable claims from text, obtain fast verdicts from multiple models, and conduct follow-up inquiries grounded in verification. This robust framework is suitable for various applications, from incident triage to pre-release continuous integration, making it a valuable resource for AI developers and users alike in verifying the accuracy of their claims.