Are we sure we're losing jobs to AI? (www.businessinsider.com)

Since ChatGPT’s 2022 debut, white‑collar job fears have stalked offices—and recent mass layoffs at big tech have sharpened that anxiety. Employers such as Amazon (14,000 roles cut), Microsoft (9,000 earlier this year) and IBM (saying it replaced hundreds of workers with AI) have explicitly cited AI-driven efficiency as part of their rationale. That makes it plausible that automation is starting to displace roles previously thought safe. At the same time, evidence is mixed: an MIT study finds roughly 95% of corporate AI investments so far deliver “zero return,” and companies routinely choose simple narratives when announcing cuts. Herd behavior, PR framing, and long-standing cost pressures complicate the claim that AI is the primary culprit. For the AI/ML community this moment matters for two reasons. First, it highlights the urgent need to demonstrate real, measurable ROI from deployed models—beyond proofs of concept—using rigorous evaluation, monitoring and human‑in‑the‑loop safeguards. Second, it underscores policy and product implications: transparent reporting on automation’s labor impacts, responsible deployment practices, and investment in reskilling are now central to trust and adoption. Bottom line: AI is being used to automate tasks and justify headcount changes in some cases, but distinguishing genuine automation-driven displacement from convenient corporate storytelling requires better data, accountability and deployment‑level evidence.