Antitrust and competition-law challenge to AI market concentration

Frame the AI-good demand through antitrust and competition-law argumentation — filing merger challenges (Microsoft's OpenAI stake, Google's Character.AI absorption, Amazon's Anthropic investment), Section-2 monopolisation cases against vertically integrated cloud-compute-plus-frontier-model incumbents, DG-COMP investigations under Article 102 TFEU, Digital-Markets-Act gatekeeper designations, CMA market-investigation references, state Attorney-General cases under state competition law. The vehicle is civil-society amicus briefs, coalition submissions to competition-authority consultations, direct legal complaints to enforcement bodies, and the sustained scholarly-and-advocacy infrastructure that supplies competition regulators with the technical and market-shape evidence they need to act.

An actor chooses this strategy because the AI industry is one of the most concentrated in modern capitalism — a handful of vertically-integrated firms hold the compute, models, and cloud-distribution channel simultaneously — and competition-law jurisdiction already exists in most major economies without new legislation. It also reframes the AI-good question in terms competition-law culture already recognises: market power, foreclosure, tying, self-preferencing, and vertical-integration harms map cleanly onto the antitrust case-book and give the movement a rhetorical entry into economic-policy venues (Federal Trade Commission enforcement, House Judiciary Antitrust Subcommittee oversight, EU Commission proceedings) it does not otherwise reach. Successful cases can force structural remedies — divestiture, interoperability, non-discrimination — that no other movement strategy has produced in this corpus.

It trades speed and coalition breadth for enforcement force. Competition cases run on multi-year timelines, and the strategy's substantive win requires either an enforcement agency (whose political leadership can decline the case) or a class of plaintiffs with standing and the resources to litigate against the biggest firms in the economy. The strategy is also structurally uneasy inside the AI-good coalition: some movement partners argue that concentration is the safety case (fewer labs, easier governance) and read antitrust as accelerationist; some argue that de-concentration would spread the same harms across more firms and worsen the accountability surface. The intra-movement disagreement is real and constrains the coalition-lobbying arm's ability to speak with one voice.

This strategy differs from strat-class-action-litigation-against-private-ai by cause of action — class-action fires on defined harms to defined plaintiffs, antitrust fires on market-shape harms to consumer welfare or the competitive process — and from strat-strategic-litigation-against-algorithmic-state-decisions by target (private firms in a market vs. public bodies in a decision). It pairs with strat-public-option-ai-infrastructure as the market-shape complement to a public-provisioning demand — antitrust breaks the incumbents, public option builds an alternative — and the two arms together are the corpus's strongest posture on the industrial-structure question the movement rarely engages directly.