University of Pittsburgh study finds AI-driven layoffs reduce employee AI sentiment and firm productivity, offsetting expected efficiency gains

Research from the University of Pittsburgh, analyzing millions of Glassdoor reviews, thousands of corporate financial reports, and hundreds of AI investment and layoff announcements from US public companies over five years, finds that AI-linked layoffs generate negative employee sentiment toward AI, which in turn measurably suppresses firm-level productivity outcomes. The study identifies job-security fears as the dominant driver of anti-AI sentiment among workers, while management sentiment on earnings calls is consistently optimistic but shows no statistically significant relationship to productivity results. For financial institutions, this creates a structural wedge between announced AI investment rationale and realized productivity returns, complicating the link between AI capex and near-term profitability improvement at client firms.