Singapore-based Sparkonomy, founded by three former Google executives, launched the Open Creator Graph on September 29, 2026. The free public platform is designed to give creators control over how artificial intelligence systems use their identity, addressing concerns about unauthorized training on faces, voices, and likenesses. The launch includes 1 million creator records available to be claimed worldwide, with no fees or follower minimums.
AI Misrepresentation Study Highlights the Problem
The launch follows Sparkonomy's AI Misrepresentation Study, which tested ChatGPT, Gemini, Grok, Muse, and Qwen using 50,000 queries about 10,000 creators worldwide. The study found that only 1 out of 10,000 creators was accurately described by all five systems, while 67% received no accurate description from any of them. Even among creators with more than 5 million followers, 37% received no accurate description from any AI engine tested.
How the Open Creator Graph Works
Creators can claim their records by signing in with an existing account from Instagram, YouTube, TikTok, or X. Within minutes, they receive a single cross-platform verified record that any AI system can read, making previously invisible creators reliably discoverable. The record also carries AI consent terms for attributes such as name, image, likeness, voice, movement, and signature.
For each identity attribute, creators choose one of three levels: Allowed, Conditional, or Prohibited. Allowed permits any use, while Conditional allows AI to show the creator accurately but requires permission before creating new content such as deepfakes or voice clones. Prohibited prevents the creator from being shown, trained on, or generated, and attributes without a set term are marked unknown rather than assumed as permission.
Machine-Readable Records and Legal Evidence
Every record is published in human-readable HTML and open machine-readable formats including JSON-LD, XML, TDMRep, and plaintext. This approach aligns with EU copyright law and the EU AI Act, which require rights reservations for online content to be machine-readable. Consent certificates are digitally signed, added to a tamper-evident public log, and independently timestamped under the EU's eIDAS framework.
Moving from Paper Rights to Practical Rights
Across most of the world, a creator's silence is often treated as permission, placing the burden of proof on creators after their identity has already been used. Courts in India have granted protection over name, voice, and likeness, while the United States has advanced Tennessee's ELVIS Act and the federal NO FAKES Act. The Open Creator Graph aims to provide a public, timestamped, machine-readable way for creators to declare their terms in advance and preserve that declaration as evidence.
Addressing Visibility and Consent Together
Sparkonomy treats visibility and consent as two parts of the same problem rather than separate issues. A creator that AI cannot find loses work opportunities, while a creator whose identity is used without permission loses control over their own brand. Most existing efforts address only one side, but the Open Creator Graph allows creators to be found and set terms in the same place.
Founders Emphasize Creator Control
Guneet Singh, co-founder of Sparkonomy, said AI operates at machine scale while consent still operates at human scale, which is why the company built a free public record for every creator. Megha Thareja Tyagi added that the mission is to spark creator livelihoods by giving creators visibility with AI and control over how it uses them. Vipasha Joshi noted that creators of all sizes want a simple way to be found and to say what is off-limits.
Sparkonomy's Open Creator Graph represents an attempt to close the gap between the rights creators hold on paper and the practical tools available to exercise them. By combining identity verification, consent management, and machine-readable evidence in one free platform, it addresses discovery and permission as a single challenge. As AI systems continue to reshape how creative work is found and used, the demand for such practical, creator-led infrastructure is likely to grow.