Mercor’s CEO Brendan Foody has criticized Scale AI for losing focus on product quality, despite its strong sales and distribution capabilities. Speaking on the "20VC" podcast, Foody praised Scale AI’s former CEO Alexandr Wang as a standout leader in growth and outreach but said the company drifted from its core mission of delivering high-quality data at scale. “In some ways, Scale lost the focus on product, on scaling quality,” Foody said. “And that was one of the largest challenges of the business.” Scale AI responded through a spokesperson, Joe Osborne, who pushed back on the criticism, stating that Mercor’s comments may be motivated by publicity. Osborne emphasized that Scale’s data quality metrics are at record highs and that the company remains committed to leading the market. The dispute comes amid a period of upheaval at Scale AI. In June, the company secured a $14.3 billion investment from Meta, valuing it at $29 billion. The deal sparked concerns about data security and independence, especially after Business Insider revealed that Scale had used public Google Docs—some containing confidential information from clients like Google, Meta, and xAI—to track work. Scale acknowledged the issue, saying it had since secured access to the documents. In July, Scale AI laid off about 200 full-time employees—around 14% of its workforce—and 500 contractors. Interim CEO Jason Droege admitted in an internal email that the company had become overly bureaucratic, with redundant layers and unclear mission alignment. Meanwhile, Mercor, founded in 2023, is positioning itself as a quality-focused alternative. Foody highlighted that Mercor pays significantly more—$95 per hour on average—compared to Scale’s reported rates of $30 to $50 for STEM experts and $15 to $30 for generalists. Mercor claims to recruit elite annotators, including International Math Olympiad medalists, Rhodes Scholars, and Ph.D. students. The contrast underscores a growing divide in the data annotation industry: while Scale AI has scaled rapidly through partnerships and capital, rivals like Mercor are betting on premium talent and higher pay to deliver superior training data. As the AI race intensifies, the debate over quality, cost, and trust is becoming central to who will power the next generation of models.