Imposing enormous damages awards may incentivize some defendant companies to voluntarily redesign litigation-spawning products and services. If changing some design elements can prevent plaintiff victories without compromising the quality, value, and consumer appeal of the goods in question, then why not implement them? It’s akin to the theory that punitive damages awards will “deter similar acts in the future.”

But what if a court goes beyond making a company pay hundreds of millions of dollars and additionally compels it to change the design of its speech-based services? I’m not talking here about Meta’s recent, voluntary multistate settlement, but rather two related decisions in New Mexico against Meta—one by a jury in March for $375 million, the other by a judge in August for another $567 million—that illustrate both situations.

Forcing Meta to pay a combined $942 million (on top of damage awards and settlements elsewhere, plus “an avalanche” of pending lawsuits) might incentivize it to voluntarily change—beyond steps it already has taken—allegedly addictive and harmful design features on Instagram and Facebook. But that’s ultimately a business and engineering decision. It’s very different from a judge mandating such modifications.

The $375 million award stems from a jury’s March 24 determination that Meta violated a state law by engaging in unfair, deceptive, and unconscionable trade practices. Among other things, New Mexico alleged that these “acts and practices . . . duped young users, their families, its own advertisers, and the public regarding the safety of their platforms and Meta’s efforts in prioritizing well-being.”

The August decision, however, involves more than just money. Based on a public nuisance theory, it takes the damages-may-incentivize-changes business equation a troublesome step further, with a judge ordering Meta to make design alterations for Instagram and Facebook. Key mandated modifications, which apply to users under age 18, affect the timing of push notifications, the display of “like” counts, and the maximum time per month the platforms can be used. Additionally, Meta must beef up its age-verification efforts and bar its chatbots from having sexualized, romantic interactions with minors.

To Chief District Judge Bryan Biedscheid’s credit, the changes aren’t as drastic as they might have been. They don’t affect Meta’s recommendation algorithms or infinite scroll and autoplay design features, partly because Biedscheid wisely recognized both First Amendment speech concerns and Section 230 immunity issues affecting them. Nonetheless, the decision illustrates problems with stretching public nuisance laws to apply to speech-based services and allowing judges to abate nuisances through design mandates.

The compelled changes are in addition to Biedscheid ordering Meta to pay $567 million to fund efforts to abate a public nuisance. He concluded that Instagram and Facebook cause and significantly contribute to an ongoing “youth mental health crisis” in the Land of Enchantment that psychologically harms minors and risks their online sexual exploitation.

The mental health crisis, Biedscheid determined, affects “public schools, hospitals, and law enforcement,” as well as minors’ families and communities, thereby rendering it a “public” nuisance. The $567 million is primarily designated for (1) treatment; (2) screening and assessment; and (3) education, awareness and prevention efforts. [Meta is appealing both the jury verdict and Biedscheid’s ruling.]

So, what’s wrong with judges forcing platform design changes? First, as Techdirt’s Mike Masnick writes, Biedscheid appointed “himself the product manager for Meta’s social media products.” That’s problematic because:

Once we’re in a world where a judge gets to make product decisions, bad things are going to happen. Judges are not product designers. Judges are not trust & safety experts. Judges are not child safety experts. They do not understand how all these things work together. They do not understand the tradeoffs of their decisions.

As Masnick rhetorically queries, “even if you hate Facebook and think it’s harmful and you trust this judge, do you trust all judges out there to get this right?”

Second, the non-monetary abatement measures raise important separation-of-powers questions regarding who—judges or lawmakers—should compel any design changes that affect speech-centric services featuring (and allowing users to engage in) content that’s generally protected by the First Amendment. Biedscheid acknowledged this issue, noting that “regulation of industry-wide features that contribute to problematic use, such as infinite scroll and autoplay, require legislative or executive branch action.”

Third, one can envision states broadly deploying public nuisance theories against generative artificial intelligence companies and ordering design changes affecting their speech tools. States may allege that GenAI tools carry “cognitive costs” and can harm students’ critical thinking skills, cognitive development, and “emotional well-being, including their ability to form relationships, recover from setbacks, and maintain mental health.” This, they might assert, affects public schools, public mental health resources, and even a state’s future workforce, thereby constituting a nuisance requiring abatement through design modifications.

Beware court-ordered, public nuisance abatement measures affecting speech technologies.