Premium, Regular, or Collusive? Brazil’s Aprix Case Tests Algorithmic Pricing

Few antitrust investigations find their theory of harm laid out in the target’s sales brochure. Brazil’s investigation of Aprix, a startup that sells pricing software to gas stations, nearly managed the feat. One of the company’s promotional brochures introduced prospective clients to the prisoner’s dilemma, the classic game-theory example in which individually rational choices can leave everyone worse off. It explained how a price war could produce just that result for rival stations, then posed the sales pitch as a question: “Which pricing decision increases the company’s profit with the lowest risk that the whole market ends up earning less? It is to answer this question that our pricing technology exists.”

Aprix’s webinars struck the same theme. They urged clients not to chase sales volume through discounts: “Lowering price almost never pays off. Resist!” A gas station that “attacks the market,” Aprix warned, would invite retaliation until “everyone loses together.”

The brochure and a handful of promotional videos helped prompt Brazil’s leading algorithmic-pricing case. In April 2026, the Tribunal of the Administrative Council for Economic Defense (CADE) approved a settlement (known in Portuguese as a Termo de Compromisso de Cessação (TCC)) with Aprix. The case produced neither a litigated finding of collusion nor a condemnation on the merits. Still, the settlement offers a concrete guide to how CADE may handle future algorithmic-pricing cases. Aprix is not CADE’s only such case, and Brazil’s debate over whether pricing algorithms can facilitate collusion remains far from settled.

Read the full piece here.

The Federal Communications Commission’s (FCC) Build America Agenda rests on a simple premise: Stop making it so hard to build. Federal and state policymakers have spent years reducing permitting delays, resolving pole-attachment disputes, and easing access to public rights-of-way—the public corridors used for infrastructure such as roads, utility poles, and fiber lines. These obstacles slow infrastructure deployment and raise its cost. Yet even as the federal government removes regulatory barriers, local governments often march briskly in the opposite direction.

That is the story unfolding in Maple Grove, Minnesota. The city has told a fiber provider that it cannot use public rights-of-way until it signs a cable television franchise, the local authorization traditionally required to operate a cable system. Never mind that the company provides broadband over fiber, not cable television.

The city’s demand is almost certainly unlawful. By imposing these requirements to raise revenue, Maple Grove is undercutting federal policy and making it harder for residents to obtain high-speed broadband.

Read the full piece here.

Antitrust law’s usual instruction to competitors who agree on price is admirably brief: Don’t. Labor law makes a deliberate exception for employees who bargain collectively. Advocates now want that exception to cover at least some independent contractors, including rideshare drivers, truck owner-operators, consultants, and other small-business owners.

The proposal may sound like a tidy way to counter the power of large platforms. It raises much messier questions. Who qualifies for the exemption? Who represents contractors with different interests? And what happens to prices, output, market entry, and workers who prefer flexible or individually negotiated terms? Answering those questions requires examining the economics of collective bargaining, the legal and political difficulty of defining a new exemption, and the costs that protected coordination may impose on workers outside the bargaining group.

The labor-antitrust exemption reflects a durable political compromise. It should not become a blueprint for shielding collective price-setting by independent contractors from competition. A better approach would preserve competition and flexible work while addressing specific worker-welfare problems through portable benefits, clear classification rules, and fewer regulatory barriers to entry and mobility.

Read the full piece here.

dvertising’s oldest rule is simple: Know your audience. In California, following it may get an online platform sued. Show retirement-home ads to seniors and student discounts to college students, and sensible targeting can suddenly look like unlawful discrimination.

That question is now before California’s Sixth Appellate District. The dispute arises from a trial court ruling in Haynie v. Google that could turn routine age-based ad targeting into a violation of California’s Unruh Civil Rights Act. The International Center for Law & Economics (ICLE) filed an Aug. 17 amicus letter supporting Google’s petition for a writ of mandate in Google LLC v. Superior Court.

If the ruling stands, it could cause widespread, unintended harm to consumers, small businesses, and the digital economy. Multisided platforms—online services that connect businesses with users—would likely restrict advertisers’ use of age data. The predictable result would be less relevant and less age-appropriate advertising for everyone.

Read the full piece here.

The whole country ostensibly wants America to win the artificial intelligence (AI) race. A striking number, however, would prefer someone else’s town to host the data centers, power plants, transmission lines, and cooling systems required to run it.

Adam Smith knew the type. In “The Theory of Moral Sentiments,” he warned against the “man of system,” who imagines society as a chessboard and believes he can move human beings as easily as pieces.

Today’s man of system has a data-center plan. Governors, legislators, regulators, and activists increasingly speak as though they can determine where enormous new electricity loads will locate, which power sources will serve them, how their owners will bargain with utilities, what labor terms they will accept, and how much support they will provide their communities—all while preserving low rates, grid reliability, environmental goals, and America’s lead in AI.

The chess pieces, it turns out, have a motion of their own.

In Pennsylvania, Gov. Josh Shapiro unveiled the Governor’s Responsible Infrastructure Development Standards, or GRID Standards, as the terms developers must meet to receive faster permitting, tax incentives, and coordinated state support. The Pennsylvania House voted 134-68 in June to codify them. When the Senate did not act, Shapiro told the Pittsburgh Business Times that he would consider executive action.

In Texas, Gov. Greg Abbott ordered regulators earlier this month to freeze approvals for data centers seeking grid connections until the state completes an audit. New York has imposed a one-year statewide moratorium on permits for the largest facilities.

The political pressure is clear. A July Quinnipiac poll found that 74% of Pennsylvania voters opposed an AI data center in their community. Republicans, Democrats, and independents agreed. The New York Times called the opposition perhaps “the most bipartisan issue since beer.”

Some of that resistance is exaggerated, emotional, and plainly hostile to growth. Other objections deserve a serious answer. Data centers are large industrial facilities, and residents are entitled to ask about noise, land use, water, air emissions, electricity bills, tax abatements, and the integrity of local decision-making. The secrecy surrounding some projects has deepened public distrust. A credible free-market case for data centers must acknowledge legitimate costs and concede that hyperscalers—companies that operate enormous networks of data centers—do not always strike defensible bargains.

Yet policymakers commit a grave error when they treat a project’s costs as grounds to stop an industry. That response treats scarcity as evidence of market failure, assumes public officials can identify the correct technical response in advance, and interrupts the decentralized adjustments already underway.

It also mistakes the visible building for the demand it serves. Demand for cloud computing, cybersecurity, medical research, financial services, logistics, streaming, and AI persists after a government prohibits a data center. The facility simply goes elsewhere and takes its investment, infrastructure, tax base, and accumulated knowledge with it.

Austrian economics focuses on how dispersed knowledge, prices, and entrepreneurial experimentation help people adapt to scarcity. Through that lens, the relevant question is whether moratoria and prescriptive mandates improve the process by which firms, utilities, communities, and consumers reconcile rapidly growing demand with limited supplies of electricity, water, land, labor, and capital. In reality, such policies obstruct that process.

Read the full piece here.

Nigeria has looked south and seen a $40 million payday for the press. The trouble is that it misread both the price tag and the fine print—and its attempt to collect may leave Nigerian publishers with fewer readers and no comparable payday.

On July 6, Nigeria’s Federal Competition and Consumer Protection Commission (FCCPC) announced investigations into Meta, Alphabet, X, and unnamed generative artificial intelligence (AI) companies. The announcement followed a petition to the Nigerian presidency from the Nigerian Press Organization (NPO).

The FCCPC identified three concerns: market dominance, the use of copyrighted news content to train AI models, and the absence of “equitable commercial engagement.”

The press release’s final paragraph makes the FCCPC’s model explicit. It claims that a similar inquiry in South Africa ended with Google agreeing to pay South African news organizations 688 million rand ($40 million) annually for three to five years. As I explained in a previous piece, regulatory ambition has its own politics: An agency’s next move often follows the path laid by its counterparts abroad.

But the FCCPC has misread the South African precedent in two important respects. First, 688 million rand is the total Google committed over five years, not an annual payment. Second, Google negotiated that payment under a statutory market-inquiry regime, a formal process that may give the South African Competition Commission (SACC) real power to impose remedies. Nigeria has no comparable regime.

Nigeria is importing South Africa’s answer without South Africa’s legal machinery.

Read the full piece here.

The AI industry has entered its manifesto era. Executives, researchers, and employees are issuing rival plans to keep advanced models safe. The fine print contains a less advertised question: Would those plans protect the public—or protect today’s leaders from the open models gaining on them?

That competition question starts with open-source AI models. These models make their source code, training methods, and trained parameters publicly available under a permissive license, allowing others to use and modify them. As Dirk Auer and I previously wrote, citing Susan Athey, then-chief antitrust economist at the U.S. Department of Justice (DOJ), a few strong open models may be enough to constrain proprietary large language models (LLMs):

… it is important not to neglect the role that open-source models currently play in fostering innovation and competition. As former DOJ Chief Antitrust Economist Susan Athey pointed out in a recent interview, the AI industry “may be very concentrated, but if you have two or three high quality — and we have to find out what that means, but high enough quality — open models, then that could be enough to constrain the for-profit LLMs.” Open-source models are important because they allow innovative startups to build upon models already trained on large datasets—therefore entering the market without incurring that initial cost. Apparently, there is no lack of open-source models, since companies like xAI, Meta, and Google offer their AI models for free…

The same reasoning applies to open-weight models. These models make their trained numerical parameters, or “weights,” freely available for download, even if their training data, full code, and methods remain private. The weights encode what a model has learned during training.

Open-weight models allow startups to build AI products and services without bearing the substantial upfront training costs that well-resourced incumbents can more readily absorb. Open-source and open-weight models therefore offer one reason, among others, for optimism about competition in AI markets.

That competitive role has made open-weight models a central target in the industry’s new manifesto war. Legitimate concerns about safety and potentially illegal conduct have prompted several private-sector regulatory proposals. The most concrete concern involves Chinese developers’ alleged large-scale distillation of proprietary models, a process in which one model learns from another model’s outputs. The proposals range from mandatory approval before release to targeted restrictions on open-weight and open-source development.

As Kristian Stout has argued, such restrictions are more likely to create problems than solve them. This post examines the competitive consequences of the most prominent proposals. Several could distort an AI market that has proved more open and competitive than the prevailing regulatory narrative suggests. The manifestos promise safer AI. Their fine print may promise today’s leaders a safer market.

Read the full piece here.

Since the COVID-19 pandemic and the political upheavals that followed, concerns have grown that many Western democracies face authoritarian pressure and are becoming less democratic. This shift has renewed interest in one of political science’s most influential frameworks for understanding democratic and nondemocratic rule: Selectorate Theory, developed by Bruce Bueno de Mesquita, Alastair Smith, Randolph M. Siverson, and James D. Morrow in “The Logic of Political Survival” (2003).

The theory fits within the law & economics tradition because it offers an economic account of political behavior. It focuses on the incentives facing every leader and treats public law as endogenous to those incentives—that is, as the product of survival strategies operating within a given institutional framework.

Bueno de Mesquita et al.’s formal model has broad implications for public finance, yet they do not develop those implications into a positive fiscal theory, meaning a theory that explains how fiscal policy works rather than how it should work. This piece takes up that task, analyzing tax and spending laws as survival mechanisms that determine who is taxed and who receives benefits to keep the leader in power.

Read the full piece here.

Australia’s latest plan to make digital platforms pay for journalism has an unusual feature. A platform can owe money even if it carries no journalism at all. The government calls this an “incentive.”

On Aug. 3, the Australian government finalized legislation establishing the News Bargaining Incentive (NBI). The government first proposed the NBI in December 2024, then put it on hold after calling a general election a few weeks later.

The NBI seeks to encourage digital platforms to enter into commercial agreements supporting Australian journalism. Large platforms that decline to enter into or renew qualifying agreements must pay a charge based on advertising revenue attributable to the Australian market. Platforms with qualifying agreements may reduce their liability through a nonrefundable offset.

The NBI’s unusual reach is deliberate. It applies to large platforms that provide significant social-media or search services, regardless of whether they carry news content. According to the government, this approach closes a major gap in Australia’s existing bargaining code, which allows platforms to avoid payment obligations by removing news.

The government has openly expressed frustration that the code has failed to secure the desired transfer of revenue from large platforms to news publishers, despite its mandatory bargaining process and binding final-offer arbitration. The NBI responds by compelling payment even when a platform makes no use of news content.

This latest change to Australia’s legal framework warrants attention for two main reasons.

Read the full piece here.