CopySight Announces $3 Million Seed Round and CopyScore V2

Citybiz reports that CopyScore V2 evaluates AI-generated video frame by frame, extending functionality previously used for images. FinSMEs describes the system as assessing exposure across trademarks, fictional characters, brand designs, celebrity likenesses, and art styles, while generating an immutable chain-of-creation log for compliance documentation.

Dealroom similarly reports that the platform compares characters, faces, logos, and styles with reference libraries to flag possible IP risks. The retrieved reporting does not detail the underlying models, reference-data provenance, false-positive rates, or the legal standards used to determine a risk score. Those details are material for teams evaluating whether automated screening can be incorporated into production approval processes.

CopySight co-founders Artem Petrov, CEO, and Konstantin Orlov, CTO, lead the company, according to Citybiz and FinSMEs. FinSMEs lists AGBO, ArentFox Schiff, and OpenArt among its clients.

CopySight says it has processed more than 87,000 copyright and IP risk assessments since January 2026, a 25-fold increase in usage over that period, according to Citybiz. The reported workflows span studios, technology platforms, legal teams, and enterprise customers.

Mucker Capital partner David Borcsok described the company as addressing capabilities related to risk and ownership in generative AI content creation. "AI cannot scale without trust," Borcsok told Citybiz. "They are addressing the essential capabilities around risk and ownership that will only become more critical as AI evolves."

The funding arrives as generative-video systems increase the volume of assets that creative, marketing, and media organizations may need to review before commercial distribution. Companies deploying comparable screening systems typically face a distinction between technical similarity detection and legal clearance: a model can flag resemblance or potential exposure, while ownership, fair use, licensing scope, and jurisdiction-specific rights often require human legal assessment.

For ML and platform teams, the operational questions extend beyond a risk score. Useful governance integrations generally require auditable inputs and outputs, reproducible model and database versions, escalation workflows, and clear retention controls for the content submitted for analysis. The available reporting indicates that CopySight records a chain of creation, but it does not specify interoperability, API behavior, or independent validation of its scoring architecture.