Franchises rarely improve by the third film when the first two already flopped with audiences. Yet that is essentially what California lawmakers are attempting with the BASED Act: a Sacramento-set third installment in an antitrust cinematic universe that began in Brussels with the European Union’s Digital Markets Act (DMA) and continued in Washington, DC with the American Innovation and Choice Online Act (AICOA).
The premise has been familiar from the start of the series. Cast a small number of large technology firms as villains. Write sweeping rules against “self-preferencing” and offering a convenient one-stop-shop with a variety of integrated tools and services. Promise a populist ending in which consumers and small businesses win by punishing bigger businesses. And, as with many overambitious films, the trailers sell something the final cut never actually delivers.
The Brussels original, the DMA, was the franchise’s expensive and self-important debut. In production since at least 2020 and released in all formats in March 2024 when the DMA became fully effective, it targeted mostly U.S.-based digital services with so-called “gatekeeper” designations that imposed bans on self-preferencing, interoperability mandates, and data-portability requirements. Supporters praised the ambitious scope. Consumers and businesses have actually been left with a clumsier user experience and a more expensive production.
The numbers are not subtle. A 5,000-person survey across 20 European countries by Nextrade Group found that two-thirds of Europeans now need more clicks or longer searches to find what they need. Among heavy search users, 61% said a single query takes significantly longer, with a weighted average of 17% longer search times. About 59% of Europeans said they would pay to restore the pre-DMA experience, with a per person willingness to pay among that 59% averaging about $300 per year. A second survey by ECIPE, conducted by Ipsos across seven EU countries, found that a 39% plurality of respondents said routine online tasks had become more cumbersome, and there was no evidence of lower prices, better privacy, or more contestable markets.
Far from a policy blockbuster, the DMA was a feature presentation that lost the audience so thoroughly the median audience member was willing to pay to leave the theater early.
The economic reviews were even harsher. A study by Cennamo, Kretschmer, Constantiou, and Garcés estimated that the DMA could cost EU service sectors up to EUR€114 billion (USD$132 billion) annually, about 0.64% of turnover, with particularly heavy losses in accommodations, hospitality, and retail. Revenue losses reached as much as €1,122 (USD$1,300) per European worker. Those losses reflect exactly the kinds of disruptions one would expect from rules that weaken search visibility, ad personalization, platform integration, and consumer trust. As the study noted, the DMA’s consent requirements caused click-through rates for targeted ads to collapse because only 5% to 14% of users opt in under the new rules, compared with roughly 90% under opt-out. That raises customer-acquisition costs for the very small businesses the regulation was supposed to help.
Meanwhile, a CCIA Research Center study published in July 2025 found that EU digital regulations impose up to $97.6 billion annually in costs and revenue losses on U.S. companies, with the DMA alone accounting for roughly $1 billion per year in direct compliance costs. That is money not spent on product development, security, or startup investment.
And then there is the supposed “Kill Zone” plot twist. Supporters of the DMA alleged that large U.S. digital services were discouraging startups by buying up potential competitors early, and a law like the DMA was needed to stop this. A CCIA Research Center study of 2,276 European venture-backed companies from 2015 through 2024 found no evidence that there was any such problem for the DMA to solve. Surprise–the purported victim was still alive and thriving! European venture activity grew steadily through 2021, with deals doubling from 7,400 to more than 15,000. But after the DMA’s passage, the seven designated gatekeepers made zero acquisitions of European venture-funded startups in 2023 and 2024. European VC investment fell 37% from 2022 to 2024, even as global stock markets rose 40%. In the denouement, the alleged Kill Zone “problem” was revealed to be fictional, while the funding drought caused by the DMA “solution” turned out to be real.
Washington then attempted a direct sequel with AICOA. Displaying all of the originality of the Force Awakens retelling the core beats of A New Hope, the setting changed, but the script did not. AICOA drew direct inspiration from the DMA, defining covered entities by size, applying special restrictions to firms with at least 50 million monthly active U.S. users and market capitalization above $550 billion. It prohibited self-preferencing in rankings and displays, restricted use of nonpublic business-user data, limited tying conduct, mandated interoperability and data portability, and offered an affirmative defense for conduct that was “narrowly tailored” and “procompetitive.”
Economic analysis was brutal. One study found AICOA would cost $319 billion without meaningfully delivering on its promised benefits, a veritable debacle of Waterworld proportions lacking even a happy ending for consumers and small businesses.
Congress screened the AICOA sequel and declined to place it in theaters. Twice. The bill lost five cosponsors when reintroduced and never reached a floor vote in either chamber.
Now comes California’s BASED Act, introduced on March 18, 2026, as an amended version of SB 1074. If the DMA was the Brussels original and AICOA the D.C. sequel, BASED is the Sacramento third entry determined to go bigger. The bill raises the threshold for covered entities to 100 million users and $1 trillion in market capitalization, but otherwise follows the same franchise formula: identify a few firms by size, then impose broad conduct rules untethered from demonstrated consumer harm.
It also adds new AI regulatory provisions, prohibiting covered entities from designing, deploying, or using AI “in a manner that systematically favors” their own products, services, or content. The bill explicitly reaches generative AI summaries, recommendation engines, conversational agents, and shopping assistants.
This is where the production goes from derivative to implausible. AICOA had no comparable AI provisions. BASED assumes digital services can prove a large language model used a “neutral methodology” rather than one influenced by commercial incentives. But modern AI systems are probabilistic. Their outputs reflect training data, fine-tuning, and reinforcement processes that are difficult to audit even for their creators. There is no widely accepted standard for certifying that an AI recommendation engine is free of self-preferencing bias. The likely effect is not cleaner competition, but legal uncertainty and pressure to scale back AI features for California users.
The bill also bans most-favored-nation clauses and margin guarantees categorically, despite economic literature showing such arrangements can lower prices under many real-world circumstances. That sort of blanket prohibition may create plenty of drama, but it is not careful policymaking.
Even California’s own Law Revision Commission seems to have rejected this genre. After more than fifteen meetings and fifty memos over several years, it concluded in January 2025 that it did not recommend a special antitrust regime for technology companies. BASED proceeds anyway.
That is the problem with this franchise. It mistakes “making life harder for big digital services” for “making consumers or small businesses better off.” The two are not synonymous. The Brussels film demonstrated that plainly. The Washington sequel failed to convince. And now Sacramento wants to greenlight a threequel with even bigger ambitions, even shakier effects, and even less evidence that audiences will leave better off.
Some franchises need a rewrite. This one needs to be canceled.
