Labor: displacement policy, algorithmic bosses, and the bargaining table

Lesson 3 of 5 in AI, Elections, Labor, Competition, and the Planet.

AI labor policy is really three separate policy problems wearing one name, and conflating them produces bad advice. The first is displacement: what happens to workers whose tasks AI absorbs. The second is algorithmic management: AI as the boss — assigning, monitoring, rating, and firing. The third is bargaining: workers negotiating AI’s terms of use directly with employers. Each has different law, different actors, and a different state of maturity.

On displacement, hold two disciplined positions at once. The economics is task-based, not job-based — models like the one behind the famous 47%-of-jobs headline were superseded by analyses showing AI automates and augments tasks within jobs, with net effects that depend on how work is reorganised. And the policy response almost everywhere is weak relative to the rhetoric: reskilling funds, education reform, and wage-insurance pilots, while more radical instruments (robot taxes, automation levies) have been debated and near-universally rejected as taxing productivity itself. The honest expert framing for a board: nobody credibly knows net employment effects; what is knowable is your workforce transition plan, and regulators increasingly ask to see one.

Algorithmic management is where hard law actually exists. The EU’s Platform Work Directive (2024/2831, transposition due December 2026) is the world’s first statute aimed squarely at the algorithmic boss: platforms must disclose the automated monitoring and decision systems they use; certain data is off-limits entirely (emotional state, private conversations, predicted union activity); and significant decisions — including terminating a worker’s account — require effective human review. It pairs this with a rebuttable presumption of employment status for platform workers. Germany adds an older instrument with new teeth: works-council codetermination, which conditions workplace monitoring technology on worker-representative consent. The US has no federal counterpart — an NLRB general-counsel memo warning about surveillance that chills organising, California’s warehouse-quota disclosure law, and a rolling set of state "no robo boss" bills whose status you should check before citing.

The third arena moved fastest of all, and it moved privately. The 2023 Writers Guild strike produced the first landmark AI collective agreement: AI cannot write or rewrite literary material, AI output is not source material that cuts writers’ credit or pay, studios must disclose when they hand writers AI-generated material, and writers’ consent governs their own use of the tools. SAG-AFTRA won consent-and-compensation rules for digital replicas of performers. These contracts govern real AI use today with a precision no statute matches — a reminder that labor law’s oldest technology, the collective agreement, is a fully operational AI-governance instrument.

Key terms: algorithmic management, platform work directive, codetermination, collective bargaining, digital replica

Interactive checkpoint quiz (2 questions) — open this page in a browser to take it.