Full to Bursting
A small newsletter about legaltech
AI
Data labelers and other workers helping to train (and in some instances, pretend to be) AI are organizing in Nairobi for better working conditions and pay [404 Media ($) | Data Workers’ Inquiry (content warning - includes descriptions of working with pornography and sexting conversations)]
From Tech Policy Press: “The Indian state of Maharashtra is developing an AI tool that uses ‘accent, tone and word choices’ to identify and deport Bangladeshi Muslims and displaced Rohingyas from Myanmar”
Two candidates for Columbia’s parliament ran for office on behalf of an AI avatar and digital platform meant to represent indigenous communities, but lost their elections [Rest of World | CBS19 News]
From Rest of World: “Black-box AI and cheap drones are outpacing global rules of war”
A New York State Senate bill would, if passed, impose “liability for damages caused by a chatbot impersonating certain licensed professionals”
Anthropic, the developer of the GenAI system Claude, is suing the Department of Defense after the agency designated Anthropic as a “supply chain risk,” as well as petitioning for review to the D.C. Circuit [Reuters ($) | TechCrunch | N.D. Cal docket | D.C. Cir. docket]
A federal judge has granted Amazon a preliminary injunction to block Perplexity from using its shopping AI agent to buy items on behalf of customers [ Marketing Brew | Reuters ($) | Docket | Order]
The developer of Grammarly, AI-based writing assistant software, is being sued after unveiling a new feature the provided AI-generated writing advice that cited the names of prominent journalists and writers without asking permission [Engadget | Platformer | Wired | PRF Law (plaintiffs’ attorney) | Docket | Complaint]
Fabrication Follies
Deutsche Bank National Trust Company v. Jean LeTennier - New York Supreme Court, Appellate Division [Above the Law | Oral Argument video | Opinion and Order]
Torrez Campus v. Munoz - Cal. Court of Appeal [The Volokh Conspiracy | Docket | Opinion]
An Extra-Special Fabrication Folly:
The fallout over AI-generated fabricated quotes in multiple filings by an AUSA in Fivehouse v. DoD continued on 3/10 with a hearing held over the judge’s show cause order; the AUSA is resigning, and the judge will issue a written opinion [ABA Journal ($) | Bloomberg Law ($) | Bluesky thread re: hearing by @randyhermanlaw.com | The Volokh Conspiracy | Docket | Order to Show Cause]
Data Privacy
An investigation by two Swedish newspapers found that Meta’s Ray Bans AI glasses are sending images and videos to data annotators in Nairobi to train AI, including content capturing people underdressing, using the bathroom and other intimate activity [BBC | Svensa Dagbladet (english) | Gizmodo | Mashable]
Related: There’s now a class action suit against Meta - Bartone v Meta [Bloomberg Law ($) | Courthouse News | Docket | Complaint]
Miscellaneous
UK users of Spotify must now submit to biometric face scans to retain their accounts, after the enactment of the Online Safety Act by the British Parliament [404 Media ($) | TechRadar]
Predictions market platform Kalshi has refused to pay out to users who predicted that Iranian leader Aytollah Ali Khamenei would be removed from power, due to a provision that does not allow event contracts that are ““directly tied to death”; as a result, users are suing Kalshi in federal court [The Independent (UK) | Reuters ($) | Docket - Risch et al v. KalshiEX LLC | Complaint]
Related: Two Congresspeople have announced that they are working on a draft of a bill to regulate prediction markets
Cryptocurrency exchange Binance (whose founder was convicted of violating anti-money-laundering statutes, then pardoned ) is suing the Wall Street Journal for defamation, as the company is the subject of probes by the DOJ, SEC, as well as a Congressional inquiry [Ars Technica | Sen. Blumenthal letter | Docket | Complaint]
Long Reads
Batista, Robert M. and Griffiths, Thomas L., A Rational Analysis of the Effects of Sycophantic AI (2026) arXiv:2602.14270
Bedard, Julie, Kropp, Matthew, Hsu, Megan, et al., When Using AI Leads to Brain Fry (March 5, 2026), Harvard Business Review https://hbr.org/2026/03/when-using-ai-leads-to-brain-fry
Perlin, Jonah, Client Confidentiality and Generative AI (February 12, 2026), 40 Harv. J.L. & Tech. __ (forthcoming 2027), Available at SSRN: https://ssrn.com/abstract=6227838
Rabanser, Stephan, Sayash Kapoor, Peter Kirgis, Kangheng Liu, Saiteja Utpala, and Arvind Narayanan. Towards a Science of AI Agent Reliability (2026) arXiv:2602.16666
