Chinese companies are conducting covert, AI-driven layoffs to circumvent government approval requirements for workforce reductions exceeding 10%. The strategy targets contractors to avoid triggering labor regulations.
Companies including major internet firms in Hangzhou have begun systematically firing contractors through quiet dismissals, sidestepping regulations that mandate government sign-off for large-scale redundancies.
China's labor laws require approval for layoffs affecting more than 10% of a workforce. By targeting contract workers individually rather than announcing company-wide cuts, firms avoid regulatory scrutiny and the formal process.
AI systems are being deployed to identify contractors for dismissal, automating performance evaluations and termination decisions. This approach allows companies to reduce headcount gradually while staying below disclosure thresholds.
The tactic reflects mounting pressure on Chinese tech firms facing economic slowdown and competitive pressures. Contract workers, who lack some protections afforded to permanent employees, are bearing the brunt of adjustments.
The practice highlights a gap in labor enforcement where companies can technically comply with regulations while circumventing their intent.
Rapid growth in artificial intelligence companies is delivering substantially higher returns for late-stage investors compared to historical benchmarks. The shift reflects accelerating valuations and market demand for AI-focused firms.
Cryptographer Bruce Schneier proposes a framework for deciding when to use AI tools: focus on whether the learning process or the final result matters most. The distinction mirrors how we approach work versus exercise.
Bluesky has expanded its AI assistant Attie beyond basic chat functions, enabling users to query trends, news, and conversations across Bluesky and other AT Protocol apps.
Anthropic's Opus 5 has claimed the #1 position on Artificial Analysis Intelligence Leaderboard, a ranking system that evaluates large language models across performance metrics.