Meta's Massive Scale AI Investment Faces Internal Strain and Rival Vendor Competition
Friction surfaces in Meta's $14.3 billion partnership with Scale AI as a key executive departs and researchers turn to competing data-labeling vendors. Internal sources report concerns over data quality within the newly formed Meta Superintelligence Labs.
Meta's massive $14.3 billion investment in data-labeling vendor Scale AI shows early signs of strain just months after the deal closes. Ruben Mayer, a top executive brought over by Scale AI CEO Alexandr Wang to help run Meta Superintelligence Labs (MSL), departs Meta after only two months. While Mayer states his departure is due to a personal matter and disputes claims about his exclusion from the core AI unit, his quick exit highlights instability in the newly integrated leadership team.
Beyond personnel changes, the technical relationship between Meta and Scale AI shifts as MSL expands its vendor network. Researchers inside Meta's core AI unit, known as TBD Labs, actively work with third-party data-labeling vendors like Mercor and Surge. These two companies rank among Scale AI's biggest competitors, making their continued use within Meta a notable snub to the newly invested partner.
The root of this friction stems from researcher dissatisfaction with the quality of Scale AI's data. Internal sources reveal that AI researchers in TBD Labs view Scale AI's output as low quality and prefer the data provided by competitors. This situation highlights a broader industry challenge, as the crowdsourcing model that initially built Scale AI's business struggles to meet the sophisticated demands of modern AI model training.