Rezolve AI has launched Rezolve Provenance, a platform designed to mark, sign and verify AI-generated and AI-modified content across text, images, voice and video as demand for identifying synthetic media increases.
The platform provides tools for AI model developers, content-generation platforms, cloud providers and enterprises to attach provenance information to content and subsequently verify its origin.
Rezolve Provenance combines C2PA Content Credentials with provenance records, verification capabilities and content-embedded watermarking for images. The system operates after content has been generated, which means companies can integrate the technology without retraining their underlying AI models.
For images, Rezolve is using watermarking technology covered by U.S. Patent No. 12,412,310. The technology encodes information into image colour channels using machine learning.
Rezolve claims the embedded signal is designed to remain detectable even when an image has been cropped, compressed, reformatted, captured through a screenshot, printed or recaptured. These resilience capabilities are based on the company's description of its technology.
The platform also includes APIs and software development kits for detection and verification, along with private-cloud deployment options. TraceWare, Rezolve's existing verification technology, provides an audit layer within the offering.
Rezolve has previously integrated TraceWare into its Brain Suite platform as part of its push towards auditable enterprise AI. In July, the company announced Auditable AI, a technology designed to allow AI-generated product recommendations to be explained and verified.
The latest launch comes as regulators and technology companies increase their focus on identifying artificially generated content.
Transparency obligations under Article 50 of the European Union's AI Act became applicable on August 2. The rules require providers of certain generative AI systems to ensure that synthetic audio, image, video and text outputs are marked in a machine-readable format and detectable as artificially generated or manipulated, subject to specified exceptions.
Rezolve's announcement also follows similar moves from major AI developers. Anthropic recently introduced machine-readable watermarking for text generated by Claude models and C2PA provenance metadata for supported files, highlighting the wider industry's efforts to establish technical methods for identifying AI-generated material.
For marketers, publishers and digital platforms, provenance technology is becoming increasingly relevant as generative AI is used to produce advertising creatives, product imagery, videos and other digital assets at scale. Machine-readable markers can provide an additional mechanism for platforms and organisations seeking to determine whether content has been generated or modified using AI.
Rezolve is initially launching a design-partner programme aimed at AI model owners, content-generation platforms, cloud providers and enterprise technology companies.
The company plans to commercialise the technology through enterprise licensing, usage-based content marking, verification API fees and private-cloud deployments.
The launch extends Rezolve's broader focus on enterprise AI verification as content provenance moves from a voluntary transparency measure towards a more formal component of AI governance and compliance.