Precautionary Antitrust: The Changing Nature of Competition Law

GAICIL Insights

Every competition law hides a forecast.¹ Precautionary Antitrust: The Changing Nature of Competition Law argues that the forecast has turned fearful. Aurelien Portuese contends that the forces that first shaped European Union antitrust, and now influence United States antitrust, are persistent and can be explained with a novel conceptual framework: antitrust has embraced the precautionary principle.

The precautionary principle is familiar from environmental and health regulation, where it licenses intervention in the face of uncertain harm. The article finds its elements in antitrust decisions and rhetoric. Imported into competition law, it justifies preemptive regulation and a static view of competition. The article's sharpest line captures the trade-off: as the precautionary principle entered antitrust, innovation exited it. Precautionary antitrust protects the current market structure, while disruptions create the next markets.

Where does this leave AI? Artificial intelligence is the hardest stress test for that approach. AI markets are young, fast and uncertain. Their future shape cannot be read from today's market shares, and the most valuable developments may come from directions no one foresees. Precaution treats that uncertainty as a reason to intervene early. The article draws the reverse inference for innovation markets: uncertainty is the natural habitat of disruptive competition, and as the status quo bias of precautionary antitrust hardens, competition grows less dynamic and less disruptive.

Here is the paradox worth sitting with. A doctrine meant to protect future competition may end up freezing present structure, and the frozen structure is exactly what disruption would otherwise have dissolved. Guarding the market against uncertainty can mean guarding it against its own renewal.

The article names what goes missing when that bias wins. Letting innovation thrive despite uncertainty requires accepting innovation defenses in antitrust cases and understanding the new business models idiosyncratic to disruptive innovators. Both, the article finds, have waned. For AI the point is concrete. Integrated offerings combining compute, models and applications, or novel licensing and partnership structures, can look suspicious through a static lens. Through a dynamic lens, they may be how innovations get financed, scaled and delivered to users.

The rhetoric around digital markets gets the same scrutiny. Large technology firms are routinely described as monopolists. The article answers that big tech firms face intense rivalry akin to monopolistic competition, and that precautionary antitrust overlooks innovation dynamics: present competition as the outcome of innovation, and innovation as a prerequisite for future competition. In AI, today's leaders compete hard with one another and with entrants for talent, compute, users and breakthroughs.

The article traces several roots of the precautionary turn, including attitudes toward risk, and its emergence first in Europe and later in the United States. That transatlantic path matters for AI governance. Rules born in one jurisdiction spread through regulatory imitation and the global reach of digital firms. A precautionary model of competition law, once entrenched, can become the template for AI markets worldwide before those markets have even matured.

None of this abandons antitrust. The article proposes principles of dynamic antitrust that foster dynamic competition, and calls for reforms: greater independence of antitrust agencies from politics, more agency staffing, truly functioning innovation defenses, and better consideration of potential competition. The goal is an enforcement regime that can spot genuine harm without crushing the disruptions that create future rivalry.

For anyone weighing an intervention in AI markets, the article supplies a two-part test. Does the measure address a demonstrated harm to the competitive process? Or does it shield the current market structure against uncertain future disruption? The first is the classic task of competition law. The second, the article argues, trades future markets for present stability.

The underlying question is philosophical and practical at once: how should a legal system treat uncertainty? Innovation is uncertain by nature. Any rule that treats uncertainty itself as a harm will tend to suppress innovation. The article explains why antitrust drifted that way and maps a route back toward a competition law that prizes the creation of new markets as highly as the protection of existing ones. In the age of AI, that choice may shape which markets exist at all.