AIImpactLab
9 impact clustersWork & marketsCognition & learningTechnical failuresPrivacySecurityEnvironmentLaw & informationSocial good & healthSystemic risk
Search the evidence

Find the signal.

Search titles, impact clusters, countries, organizations and the full text of every analysis.

3 stories found

An hourglass sits on punched time slips as a pencil points to one entry.
Law & informationUnited States+1 clusters01

The lawyers used AI. That wasn’t enough to cut their fee.

The winning side used AI. Why shouldn’t the losing side get a discount on the lawyers’ fees? The Tenth Circuit had a useful answer in Flux Drive v. Rong: show what was unreasonable about the hours. Pointing to the software did not do that work. The dispute was over fees awarded after a dismissal for lack of personal jurisdiction. It was not a client asking a lawyer to explain a private invoice. In its October 30 opinion the court upheld an initial $114,663.11 award challenged on the AI issue. It also upheld a separate supplemental award. The opinion describes time recorded in tenths of an hour, individual entries rather than blocks of undifferentiated tasks, and reductions where appropriate. The challengers did not show that the hours the lawyers actually spent were unreasonable. That record, not the mere fact of adopting technology, mattered. There are two protections here, and each side of a bill needs one. A lawyer paid by the hour cannot fairly charge for three hours of drafting that never happened. Equally, checking whether an AI tool invented a citation is work, even if generating the first draft took seconds. The ABA’s guidance on generative AI distinguishes actual time from time a lawyer hypothetically would have spent without the tool. Read together, the opinion and the guidance give clients a more useful question than “Did you use AI?” Ask what the lawyer did, why it needed doing, and what contemporaneous record supports the charge.

4 min
A European age gate closes across chatbot, social, video, and game portals while a quiet identity-verification system grows behind it.
Law & informationEuropean Union+3 clusters02

EU draft would lock under-15s out of chatbots, social media and online games

A draft European Union plan would create the bloc’s broadest age-based restrictions yet for social media, video-sharing platforms, AI chatbots, and online games. Reuters reports that the proposed EU Kids Act would allow people fifteen and older to open their own accounts. Children aged thirteen and fourteen could receive limited, parent-opened introductory accounts for social and video platforms, while accounts for ages three through twelve would be fully parent-controlled and limited to child-friendly services; children under three would have no access. The draft would also require age verification, tools for reporting harmful content, effective parental controls, and design changes intended to avoid addictive experiences and harmful feeds. Companies would pay a supervisory fee to fund enforcement. This is not law. Details can change before the announcement, and the proposal would still require negotiation with EU countries and the European Parliament. The policy’s strength is that it assigns duties to platforms rather than asking children alone to resist systems optimized for engagement. Its risk is that broad age assurance can create new identity and privacy infrastructure, while a single access rule can flatten important differences among messaging, education, play, health support, and social connection. The test should be whether the final law targets demonstrated mechanisms of harm, minimizes data collection, provides accessible appeals, and measures what children gain or lose after restriction.

7 min
A Minnesota-shaped legal seal and consent shield block synthetic image pixels from reaching a protected silhouette.
Cognition & learningMinnesota, United States+4 clusters03

Minnesota’s nudification ban shifts liability upstream to AI services

Minnesota's new law takes effect today and prohibits websites, applications, software and other services from allowing people to access, download or use nudification technology—or from generating the altered image on a user's behalf. Advertising and promotion are also prohibited. A depicted person can sue for compensatory damages, including mental anguish, plus punitive damages, legal costs and injunctive relief. The state may seek a civil penalty of up to $500,000 for each unlawful access, download or use. The law changes the burden of response: instead of asking victims to chase every synthetic image, it targets the services that make mass production possible.

3 min