Workers' Daily Calls for Building "Levees" Around AI Employment, Urging Improved Labor Standards and Algorithm Oversight
Claire Weston
China's Workers' Daily on June 11 published an editorial naming three labor-rights risks in the AI era — digital clones, AI-driven dismissals, and algorithmic surveillance — and urged faster regulation. Citi estimates 70 million Chinese jobs could ultimately be threatened; Beijing has begun warning tech firms not to use AI as cover for layoffs.
Why is the Workers' Daily speaking up now?
The Workers' Daily was founded in 1949 as "the voice of China's working class" and ranks among the country's major state-owned outlets. This means → the editorial is not just one newspaper's opinion — it is a policy-grade signal on the official labor-rights stance.
The piece calls for a labor-protection framework "fit for the digital-intelligence era," with unions and worker representatives given a seat at the table when algorithm rules are negotiated.
Bloomberg cited a Citi estimate: China's AI adoption is "broad but still shallow," yet it could ultimately threaten 70 million jobs. In plain terms = the technology has barely gone deep, and the job-displacement pressure is already in the tens of millions.
What are the three risks?
Risk one: digital-clone violations. Some firms have institutionalized building AI skill packs and training employee "digital twins" — even tying them to performance reviews. This means → companies are systematically harvesting workers' behavioral data; without proper informed consent, that crosses the personal-information red line.
Risk two: AI replacement as a layoff shield. Companies have already cited "tech upgrades" or "AI impact on projects" to unilaterally cut pay, reassign, or terminate workers. Courts have ruled clearly: adopting AI is a deliberate business choice, not an unforeseeable force majeure — dismissals on that basis are illegal.
Risk three: algorithms as "digital overseers." Opaque task-dispatch systems, arbitrarily adjusted piece rates, and unpaid "standby" time all squeeze workers' right to rest and fair compensation.
Workers' know-how is being "distilled" into AI — who owns it?
The editorial notes that the experience, methods, and tacit knowledge employees build on the job — the kind you learn by doing, not from a manual — are being fed into AI training. In plain terms = you taught the machine how to do your job, but who owns that "lesson plan" is still legally unclear.
Intellectual-property ownership sits in a legal gray zone, making labor disputes highly likely.
This reflects a deeper issue: the core asset of the AI era is not just code and compute — it is human experiential data — and the legal framework has not caught up.
What remedies does the editorial propose?
For data training: strictly enforce the Personal Information Protection Law's "informed consent" and "minimum necessity" principles — employers collecting worker data for AI must disclose the purpose and take only what is needed.
For AI-driven dismissals: require companies to fulfill prior obligations — negotiate reassignment, offer retraining. In plain terms = no firing by email; the employer must first discuss alternatives.
For algorithmic black boxes: bring algorithm rules that affect workers' rights into the scope of democratic consultation and public disclosure — how tasks are assigned and priced is something workers have a right to know and negotiate.
What does this mean for the market?
Beijing has begun warning employers, especially tech companies, not to lay off workers when adopting AI. This means → the policy signal has shifted from "encourage AI deployment" to "deploy if you want, but don't dump the social cost onto workers."
The editorial's key line: "The dividends of technological progress should be shared by all of society, not become a tool for a few employers to harm workers' rights."
This reflects the regulator's floor-setting logic: AI efficiency gains belong to companies, but the legal risks of layoffs and data violations will increasingly be pushed back onto the corporate side.
Content is for reference only, not financial advice.