Scale AI has formally announced the closure of a team of contract workers based in its Dallas office, a move that took effect on Monday and symbolizes a deliberate strategic turn toward a more technically sophisticated and expert-oriented phase of its operations, the company confirmed to Business Insider. This reorganization represents the latest in a series of internal adjustments following Meta’s acquisition of a stake in the company through a substantial $14.3 billion agreement, a deal that immediately reshaped partnerships and priorities across the rapidly evolving landscape of AI development.
The disbanded group—known internally as the New Projects Organization, or NPO—comprised slightly more than a dozen contractors who were primarily tasked with performing generalist work. Their contributions involved improving the natural language and stylistic quality of artificial intelligence chatbot responses, effectively serving as human trainers who fine-tuned the subtlety, fluency, and coherence of machine-generated text. However, this type of generalized support has become progressively less essential in the fast-transforming AI training ecosystem. As models advance in sophistication, they increasingly demand supervision and data refinement from individuals with deep, domain-specific expertise—professionals capable of providing insights rooted in specialized disciplines such as medicine, robotics, and finance. The growing necessity for precision and contextual understanding is thus prompting companies like Scale AI to redefine the skill sets they rely upon.
Headquartered in San Francisco, Scale AI’s broader mission is centered on enhancing the performance of cutting-edge chatbot systems developed by major technology companies. The firm plays a crucial role in strengthening these systems by evaluating, editing, and rewriting machine responses to ensure reliability, factual accuracy, and stylistic consistency. Within this broader context, Scale AI participates in the fast-moving field of AI training and data refinement—an area that evolves almost as swiftly as the products it helps improve.
According to an internal communication distributed by HireArt, the staffing agency that has long provided contract personnel for Scale AI, the affected Dallas-based workers were offered four weeks of severance pay and continued healthcare coverage through October as part of a transitional compensation package. Furthermore, to help ease the employment shift, Scale AI extended an invitation for the displaced team members to join Outlier, its large-scale contract and freelance platform where thousands of remote contributors assist in refining and training AI models. In this communication, the agency reiterated the company’s commitment to supporting those employees through the change process, assuring them that alternative opportunities remained available within the wider Scale AI ecosystem.
Confirming the reorganization, company spokesperson Natalia Montalvo provided a detailed statement to Business Insider. She described the Dallas group as a “small experimental onsite program staffed by a contract workforce,” adding that its closure mirrors a broader transformation across the AI sector: an undeniable trend favoring higher-level, expert-driven data work that requires analytical depth and technical specialization. Montalvo emphasized that the restructuring affected only a small portion of Scale AI’s overall workforce and that the change in no way compromises the delivery of services or ongoing client commitments. She also stressed that Scale AI is concurrently expanding its expert-level training initiatives, ensuring the firm remains aligned with the escalating technical demands of modern AI development. Importantly, she clarified that the adjustments have no bearing on other groups of contractors currently operating in Dallas, who continue to support the company’s core projects.
This strategic shift by Scale AI parallels moves taken by other AI firms such as xAI, which similarly narrowed its training operations in recent months by reducing the number of generalist AI tutors to concentrate on contributors with specialized expertise. These industry-wide changes underscore a reshaping of priorities within the AI development space, where depth of subject knowledge increasingly outweighs breadth of general capability.
The catalyst for Scale AI’s extensive internal restructuring can be traced back to Meta’s investment deal, made public in June. The partnership not only signaled a new chapter for the company but also prompted considerable uncertainty among existing clients. Shortly thereafter, several of Scale AI’s longstanding enterprise collaborators, including heavily influential organizations like Google and OpenAI, suddenly suspended joint projects. Within weeks, the company initiated a significant workforce reduction—eliminating about 14% of its total personnel, amounting to roughly 200 full-time employees and 500 contractors. In an internal message addressing the layoffs, management attributed the decision to previous overhiring patterns and unspecified market pressures, reflecting the turbulence that sometimes accompanies rapid growth in a competitive field.
Further developments occurred just last month, when Scale AI released twelve members of its Red Team—a key internal unit tasked with rigorously testing and analyzing AI models for potential risks, harm, or unintended bias. Officially, the company described those cuts as being related to performance concerns, though several former members disputed that explanation, claiming that the team’s workload had been diminishing steadily since the Meta partnership began reshaping the company’s focus. These events collectively paint a portrait of an organization undergoing significant structural and strategic realignment, seeking to sharpen its expertise and remain resilient amid the accelerating evolution of artificial intelligence technologies.
For individuals wishing to share information or insights regarding these developments, Business Insider has invited tips to be sent to reporter correspondence channels through provided secure contact options, encouraging contributors to use personal devices and nonwork networks to ensure confidentiality. This ongoing dialogue reflects the intense public and professional scrutiny surrounding the shifts occurring within one of the industry’s most prominent AI infrastructure firms.
Sourse: https://www.businessinsider.com/scale-ai-shutters-dallas-contractor-team-shift-to-expert-data-2025-10
