Even as Google spends billions to advance AI, the people hired to refine its most visible products are being dismissed. More than 200 contractors who trained and tested Google’s AI models were laid off last month, according to media reports. The job cuts were executed not directly by Google but through GlobalLogic, a Hitachi-owned outsourcing firm that manages much of Google’s AI rating work.
The affected group included many “super raters,” specialists with advanced degrees tasked with evaluating and rewriting AI responses for systems such as the Gemini chatbot and AI Overviews in Google Search. Their job was to ensure outputs sounded accurate, natural, and safe for public release. Some contractors now worry that the very systems they helped train could make their work obsolete. “I was just cut off,” said Andrew Lauzon, who joined GlobalLogic in March and was terminated by email on August 15. “I asked for a reason, and they said ramp-down on the project—whatever that means.”
This sense of being replaced by the tools they refined has added psychological weight to the layoffs. “We’re like the lifeguards on the beach—we’re there to make sure nothing bad happens,” said one rater, who compared their feedback to a safeguard engineers cannot provide on their own.
The dismissals also arrive amid efforts by some contractors to improve working conditions. Over the past year, workers pressed for better pay and more transparency around performance expectations. Several explored unionization through the Alphabet Workers Union, but contractors allege those efforts were discouraged. At least two former raters have filed complaints with the National Labor Relations Board, saying they were dismissed for speaking up. Others described an “oppressive atmosphere,” where short-term contracts and lack of benefits kept workers reluctant to raise concerns.
Pay disparities deepened the sense of unfairness. Contractors directly employed by GlobalLogic earned between $28 and $32 per hour, while those hired through third-party staffing agencies often received just $18 to $22 for identical work. Generalist raters without advanced degrees were paid even less, despite sometimes being assigned to complex tasks. Google has emphasized that it is not directly responsible for the employment status of the dismissed workers. “These individuals are employees of GlobalLogic or their subcontractors, not Alphabet,” spokesperson Courtenay Mencini said. “As the employers, GlobalLogic and their subcontractors are responsible for the employment and working conditions of their employees.”
GlobalLogic, for its part, has declined to comment publicly. This outsourcing arrangement highlights a broader pattern across the tech sector: major companies often rely on third-party firms to handle content moderation, data labeling, and other critical yet precarious tasks, thereby insulating themselves from direct accountability when disputes arise.
This development comes even as the tech titan is making investments in AI (much like other tech companies) and positioning its Gemini models and AI Overviews as cornerstones of its future strategy against rivals like OpenAI and Microsoft. Yet the very workers who made these systems usable for the public are facing instability, low pay, and sudden termination. Their role is similar to that of content moderators at social media companies — a workforce essential to the platform’s function yet often treated as expendable. Unlike full-time engineers, raters had little job security, no paid leave, and limited protections, despite being essential in giving Google’s AI products credibility and reliability.
The Tech Portal is published by Blue Box Media Private Limited. Our investors have no influence over our reporting. Read our full Ownership and Funding Disclosure →
