(Bloomberg) — Sahara AI, a startup co-founded by a University of Southern California associate professor, has raised $43 million as it seeks to address thorny copyright and privacy issues surrounding artificial intelligence technology.
Most Read from Bloomberg
The funding was led by Pantera Capital, Binance Labs and Polychain Capital and included investors such as Samsung, Matrix Partners, Nomad Capital as well as angel investors and founders from the industry, the startup said in a statement on Wednesday. It will deploy the capital to expand its team of experts and develop its technology.
The Los Angeles-based startup is building an AI platform that incorporates technology for recognizing copyrights and tracking assets. The venture aims to tackle the long-standing issue in AI of how to credit — and compensate — players involved in the technology’s development.
AI models and tools such as Google’s Gemini, Microsoft Corp.’s Copilot and OpenAI’s ChatGPT are trained on massive amounts of data and refined through user feedback, with returns largely accruing to the companies in control of the models. OpenAI and Microsoft have been sued repeatedly for their use of content, including by newspapers such as the New York Times and the Authors Guild of America.
Sahara’s platform is aimed at allowing participants to get credit for their contributions, whether for building models and tools, or providing data and expertise to make them work better. The goal is to have verified, trackable data, and then pay the person or business for their contributions, from companies that build the infrastructure to developers building tools and bots.
The company says it has about three dozen enterprise customers including Microsoft, Amazon.com Inc, Massachusetts Institute of Technology, Snap Inc. and automotive components maker Motherson Group. The platform’s data crowdsourcing has 200,000 contributors.
“We are building a collaborative AI economy,” Sean Ren, the startup’s co-founder and chief executive officer, said in a video interview. “As AI technologies evolve, it’s critical that a robust, decentralized infrastructure be available.”
Ren, a computer science associate professor at the University of Southern California’s Viterbi School of Engineering, established the company in 2023 along with co-founder Tyler Zhou, the former investment director of Binance Labs.
“The most tangible copyright application is personal data, for example, photos, text messages and emails, and we’re afraid of losing control of our personal data to the likes of OpenAI,” said Ren. “If privacy and copyright are taken care of, it could trigger a lot of exciting AI use cases such as making personal replicas of yourself.”
(Updates with new investor Binance Labs in the second paragraph)
Most Read from Bloomberg Businessweek
©2024 Bloomberg L.P.
Add Comment