Reporting this week found that Google, Microsoft, and Meta have quietly stopped publishing workforce diversity data, after a decade of embracing the practice. The timing is perfect. When you’re eliminating hundreds of thousands of jobs in the name of “AI efficiency”, transparency about who’s getting cut becomes awfully inconvenient.
American tech companies have announced close to 500,000 layoffs since 2022, with more than 89,000 workers cut in 2025 alone. Amazon eliminated 27,000 jobs; Google axed 12,000 employees whilst ramping up AI investments; Meta slashed 21,000 positions. But the layoffs haven’t hit everyone equally. Analysis of 1,157 LinkedIn profiles from workers laid off between November 2022 and January 2023 found that HR specialists and recruiters, where many DEI hires are parked, comprised 27.8% of all cuts, the largest single group. Software engineers came second at 22.1%.
When you’re planning mass cuts, you eliminate the people who could organise resistance first. HR departments and recruiters serve as organisational infrastructure which can mobilise collective action when the real cuts come, especially in America’s weak, largely non-unionised white-collar sector. By cutting DEI programmes — as one in five organisations reportedly have since Trump took office — and firing workers early under the guise of “we’re not hiring anymore,” companies neutralise the most fractious part of the workforce before the deeper automation hits.
Following the death of George Floyd in 2020, tech companies made diversity hiring pledges totalling billions. DEI job listings surged by 123%. Google promised to improve representation of underrepresented groups in leadership by 30% by 2025. Meta committed to extensive recruiting programmes. Amazon pledged sweeping changes across its massive workforce. These commitments created dedicated DEI teams, expanded HR departments, and robust recruiting functions.
All of that infrastructure is now being dismantled. Bloomberg’s analysis documented that black workers made up 26% of the 127,418 employees cut by 84 major corporations in 2023, with losses accelerating almost tenfold between 2022 and 2023. But the demographic impact is almost beside the point. Companies built expensive organisational capacity to address labour issues in 2020 when American cities burned. Now that the pressure has been lifted, they’re destroying that capacity.
Meanwhile, the consulting firm McKinsey predicts that 12 million Americans will need to switch occupations by 2030, including the thousands of McKinsey employees cut since 2023, with generative AI automating 30% of work hours. Customer service roles will decline by two million jobs; office support positions will drop by 3.7 million.
AI provides perfect plausible deniability for a wipeout that started with DEI: companies can claim they’re eliminating roles because algorithms replaced them, not because they’re targeting specific demographics or busting collective bargaining power. Salesforce has cut 4,000 customer service jobs using AI agents, reducing support costs by 17%. TikTok replaced content moderators with automation that now handles 85% of post removals. Back in March, Amazon eliminated 14,000 middle management and HR positions. Customer service, content moderation, early-career white-collar managers, administrative support: these are exactly the areas in which companies concentrated cheap, entry-level diversity hiring after 2020.
Research from the Conference Board in August found that use of the “DEI” acronym in major corporate filings dropped 68% between 2024 and 2025. Disclosure of women in management fell by 16%, while reporting on racial diversity of board directors plummeted by 31%. Google’s annual filing made no mention of diversity, despite referencing it eight times the previous year.
American businesses are now conducting historic “workforce transformations”, which they ascribe to strategy and efficiency. External pressure groups which could fight these cuts now have minimal data to rally around, and no allies inside the companies. Civil rights coalitions will find the diversity officers who took their calls have been “made redundant”. Journalists investigating demographic impacts will hit ever thicker walls of corporate opacity. Regulators will find that companies have technically violated no laws by refusing to publish voluntary disclosures.
Big Tech made extravagant and expensive “woke”-sounding promises in the summer of 2020. Five years later, they’re methodically breaking those promises while building the perfect defence. Cut the internal organisers first; stop publishing data that can help us understand the cuts; blame AI and shareholder value for everything. By the time the irreversible workforce transformation hits, there’s no one left who knows how to fight back and no union or strike action that could work. The people who used to monitor that data and occasionally serve as organisational ombudsmen are already long gone.