Meta Platforms Inc. has indefinitely suspended its partnership with AI data-training startup Mercor after the company confirmed a significant data breach linked to a supply chain attack involving the LiteLLM open-source project. The pause in collaboration, first reported by Wired, has raised concerns about security protocols surrounding sensitive AI development data.
The breach reportedly threatens critical data used in training AI models, prompting Meta to investigate the incident thoroughly. The implications of this breach extend beyond Meta, as reports indicate that OpenAI, another major player in the AI industry, is also looking into the security lapses associated with Mercor.
Sources familiar with the situation suggest that the breach jeopardizes the integrity of training data, which is vital for the development of advanced AI technologies. The halt in operations may impact not only Mercor but also other companies relying on its data services.
Despite this setback, Mercor had seen significant growth following Meta’s investment, with revenues reportedly quadrupling. As the industry grapples with the ramifications of this incident, stakeholders are closely monitoring how both Meta and OpenAI respond to the breach as well as any potential long-term effects on their projects.

