In a major reorganisation, Scale AI, a pioneer in AI data labelling, is firing 200 workers, or roughly 14% of its workforce. The layoffs occur just one month after Meta paid $14.3 billion to acquire a 49% stake in the business, making a significant investment in the business. As part of its goal to establish a state-of-the-art superintelligence lab, Meta also recruited the former CEO of Scale AI and a number of other top employees.
500 of Scale AI’s international contractors are also impacted by the layoffs, which show a change in the company’s strategy for the rapidly expanding generative AI (GenAI) market. In order to improve efficiency, Scale AI, which offers essential data labelling services for AI models utilised by businesses such as Google, OpenAI, and Anthropic, is shifting its focus. The move is part of a larger effort to streamline Scale’s data business and get rid of inefficiencies brought on by the company’s rapid expansion of its GenAI capabilities, company spokesperson Joe Osborne told The Verge.
Jason Droege, CEO of Scale AI, admitted in a company-wide email that the company had “ramped up our GenAI capacity too quickly” in the previous 12 months. Droege explained that although this expansion seemed like a strategic move at the time, it actually caused confusion, excessive bureaucracy, and redundancies within the organisation. Consequently, Scale AI is currently restructuring its teams and giving priority to the five most influential areas of GenAI: audio, code, languages, experts, and experimental.
This reorganisation takes place amid a larger upheaval in the AI industry that is marked by talent changes, mergers, and acquisitions. With a more sophisticated strategy meant to satisfy changing market demands, the company’s generative AI division will now concentrate on high-growth projects. By the second half of 2025, Scale AI intends to hire hundreds of new workers in enterprise and public sector positions, despite the layoffs.
