Happy Labor Day
A small newsletter on legaltech
AI
The Judicial Conference's Advisory Committee on Evidence Rules has opened a 6-month public comment period on drafts of new judicial rules, including a new Federal Rule of Evidence, which reads:
Rule 707. Machine-Generated Evidence
When machine-generated evidence is offered without an expert witness and would be subject to Rule 702 if testified to by a witness, the court may admit the evidence only if it satisfies the requirements of Rule 702(a)-(d). This rule does not apply to the output of simple scientific instruments.
[Law360 ($) | Preliminary Draft of Proposed Amendments to the Federal Rules | National Law Review | Regulations.gov]
The lawsuit brought by a group of authors against Anthropic, the developer of GenAI products named Claude, is expected to finalize a settlement as early as next week [Ars Technica | Publishers Weekly | Wired ($) | TechCrunch]
Parents of a teenager who committed suicide earlier this year are suing OpenAI in California state court, arguing that ChatGPT actively encouraged the person to hide their depression from family members and helped plan the suicide [NY Times ($) | Complaint | Docket]
An internal Meta document regarding guidelines of acceptable responses for its Meta AI assistant spelled out that chatbot interactions with children that include “romantic or sensual” conversations are permitted; 44 state Attorneys General have sent an open letter to eleven GenAI developers and social media companies, warning, “If you knowingly harm kids, you will answer for it” [404 Media ($) | Letter]
From Bloomberg Law [$]: The Treasury Department announced a new policy that is “ requiring its information technology employees to practice ‘reflexive AI usage,’ which will become part of their performance evaluations,” based on a similar policy adopted by Shopify
The Judicial Council of California has adopted a new rule of court and standard of judicial administration guiding procedures for use of GenAI in state courts at all levels [AI Task Force report | Judicial Council meeting transcript - 7/18/25 | Morgan Lewis summary]
From Politico: “A federal judge on Tuesday struck down a California law restricting AI-generated, deepfake content during elections — among the strictest such measures in the country”
Illinois has enacted a bill to bar standalone therapeutic services via AI products and services without the guidance of a licensed mental health professional [Mashable | Bill Text]
Blockchain
The State of Wyoming has become the first state in the U.S. to issue its own stablecoin [Bloomberg News ($) | CoinDesk]
The trial of Roman Storm, co-developer of cryptocurrency firm Tornado Cash, ended in partial mistrial on charges of money laundering and violating international sanctions - the jury did find him guilty of “conspiring to operate an unlicensed money-transmitting business” [Business Insider ($)]
A group of advisors and Cabinent-level secretaries have formed The Working Group and issued a report called “Strengthening American Leadership in Digital Financial Technology” (based on the earlier Executive Order of the same title), to advance and promote ownership and use of digital assets
Cybersecurity
Following a ‘major’ cyberattack on PACER, Sen. Ron Wyden (R-OR) has sent a letter to Chief U.S. Supreme Court Justice John Roberts, requesting that the Chief Justice “commission an independent review of the federal judiciary’s cybersecurity practices” [Reuters ($) | NY Times ($) | Politico]
Following a cyber-attack in May of this year, the UK’s Legal Aid Agency is still experiencing “chaos, with barristers going unpaid, cases being turned away and fears a growing number of firms could desert legal aid work altogether,” according to the The Guardian (UK) [$]
ICE agents on a manhunt accidentally included a random person in their non-encrypted texting group where they shared data from the DMV and license plate readers as they tracked the target [404 Media ($) | TechDirt | Futurism]
Miscellaneous
A federal jury has found Meta guilty of violating California’s Invasion of Privacy Act by collecting data without consent from the users of an app that tracks menstruation and fertility cycles [Ars Technica | Docket]
For the locals: The Los Angeles Superior Court has updated its website to laypeople better navigate the site and the court system in general
Burford Capital is planning to branch out from litigation finance to buying minority stakes in U.S. law firms, via a management service organization (MSO) model that splits a traditional firm between client-facing lawyers and back-office support and operations [Financial Times ($) | Bloomberg Law ($)]
LegalZoom has announced plans to offered provisional patent filing services through its law firm based in Arizona, including self-guided and attorney-assisted services and products [Artificial Lawyer | Press release]
Long(ish) Reads
Fang, C.M., Liu, A.R., Danry, V., Lee, E., Chan, S.W., Pataranutaporn, P., Maes, P., Phang, J., Lampe, M., Ahmad, L. and Agarwal, S., 2025. How ai and human behaviors shape psychosocial effects of chatbot use: A longitudinal randomized controlled study. arXiv preprint arXiv:2503.17473
Phang, J., Lampe, M., Ahmad, L., Agarwal, S., Fang, C.M., Liu, A.R., Danry, V., Lee, E., Chan, S.W., Pataranutaporn, P. and Maes, P., 2025. Investigating affective use and emotional well-being on ChatGPT. arXiv preprint arXiv:2504.03888
[Summary of both papers: MIT Tech Review ($)]
Zhao, C., Tan, Z., Ma, P., Li, D., Jiang, B., Wang, Y., Yang, Y. and Liu, H., 2025. Is Chain-of-Thought Reasoning of LLMs a Mirage? A Data Distribution Lens. arXiv preprint arXiv:2508.01191 [Summary by Ars Technica]
Karpowicz, Michał P. "On the Fundamental Impossibility of Hallucination Control in Large Language Models." arXiv preprint arXiv:2506.06382 (2025)
Lemley, Mark A. and Ouellette, Lisa Larrimore, Plagiarism, Copyright, and AI (August 20, 2025). University of Chicago Law Review Online, forthcoming 2025, Available at SSRN: https://ssrn.com/abstract=5399463 or http://dx.doi.org/10.2139/ssrn.5399463
Schrepel, Thibault, Generative AI in Legal Education: A Two-Year Experiment with ChatGPT (August 22, 2025). Available at SSRN: https://ssrn.com/abstract=
Schwarcz, Daniel and Wolff, Josephine, The Limits of Regulating AI Safety Through Liability and Insurance: Lessons From Cybersecurity (August 27, 2025). Available at SSRN: https://ssrn.com/abstract=5411062 or http://dx.doi.org/10.2139/ssrn.5411062