AI × Law

An international research team has launched SteuerLLM, a 28-billion-parameter AI model specifically trained on German tax law. The model demonstrates superior performance over comparable and even larger general-purpose AI systems, suggesting that domain-specific design is crucial for legal AI applications.

The EEOC has removed its AI technical assistance documents and de-prioritized disparate-impact enforcement under a 2025 executive order. However, Title VII still holds employers responsible for biased AI hiring tools, pushing exposure to private litigation, such as Mobley v. Workday, and a growing body of state laws like those in NYC, California, and Illinois.

Apple has received preliminary court approval for a $250 million settlement in a false advertising lawsuit. The suit alleged Apple misled customers by promising AI features for the iPhone 16 and some iPhone 15 models that were not available upon release. Eligible customers could receive between $25 and $95 per device, with payouts anticipated after a final approval hearing in September 2027.

A Munich court issued the first ruling on the merits of AI music training, finding that Suno infringed copyrights represented by GEMA and rejecting its fair use defense. This judgment, though expected to be appealed, sets a precedent as the industry bifurcates into licensed and unlicensed AI music, with a US fair use ruling against Suno not expected until 2027.

Spain’s tech M&A market logged 44 transactions in July 2026, including deals involving AI capabilities such as DKL, Solver Machine Learning, and a new AI compliance venture formed by Andersen España and 1MillionBot. The month also saw AI-focused venture rounds, including Sherpa.ai’s $18 million raise and Cognitfy’s $15 million Series A, signaling continued capital flow into applied AI.

Mitratech, an enterprise legal management provider, acquired Viewabill to give legal departments more real-time visibility into outside counsel time and billing information. The deal fits Mitratech’s ongoing expansion in legal tech and matters to in-house teams focused on outside counsel spend control.

A legal research comparison says Lexis+ AI and Ask Practical Law AI produced incorrect or misgrounded answers in more than 17% of queries, while Westlaw’s AI-Assisted Research was above 34%. The piece says the real shift is a hybrid workflow: lawyers use AI to map and synthesize, then still verify with Westlaw or Lexis citators and editorial sources.

Evatt AI, built by Dawnview AI Pty Ltd, says it has grown from a Munich dorm-room prototype to a legal research tool used by more than 2,500 practitioners across six jurisdictions. The company also says its verification layer uses three-model cross-checking and that it launched publicly in February 2025 with 500+ signups in its first week.

ElevenLabs posted a Commercial Counsel role in India to support SaaS, licensing, and strategic partnerships across India and, when needed, APAC. The job listing says the company has raised $781 million, is valued at $11 billion, and is building out legal support for AI-related regulatory and contract issues.

Anthropic appointed Robert Mahari as its first official Head of Claude for Legal. Mahari, a Stanford CodeX fellow and MIT legal AI PhD, will work alongside Anthropic’s legal product lead to build the company’s legal vertical for Claude. The hire shows Anthropic is formalizing its push into legal AI as rivals prepare their own offerings.

A California bill would ban companies from marketing chatbots as therapy and limit AI use in mental-health decisions without licensed-professional review. The story also notes wrongful death lawsuits against chatbot makers and a state complaint targeting Kaiser Permanente’s mental-health triage tool, making this a concrete policy and litigation development.

Docusign says its Intelligent Agreement Management platform now uses AI to analyze agreements, identify high-risk language, track renewals, and automate contract workflows. The page also highlights broad adoption, including 95% of Fortune 500 companies, making this a notable product push in legal and contract operations.

ACHNET says it has released compliance documentation for employers using its AI hiring tools, including guidance on human oversight, logging, bias testing, and recordkeeping under the EU AI Act. The pitch matters because recruitment systems are treated as high-risk, and employers using AI to screen candidates in Europe still need to meet deployer obligations even after the compliance deadline was pushed to December 2027.

The National Law Review, WashU Law, and Wickard are hosting a free virtual Legal AI Demo Day on August 11, 2026, featuring rapid demos from legal tech companies. The lineup includes products for fact management, discovery, billing, real estate due diligence, securities compliance, judicial workflows, and workers’ compensation litigation, showing where legal AI is being applied in practice.

This roundup says legal AI activity is picking up: Aavalynx raised a £1.5m pre-seed, Wordsmith raised a $14m Series B extension, GC AI launched Contract Intelligence, and Anthropic hired its first Head of Claude for Legal. For litigators and law firms, the bigger story is the push toward more secure, more capable tools that can work with firm data.

AbbVie is hiring a Senior Counsel, AI Legal – Operations & Commercial to support the operationalization and scaling of AI across the company. The role covers AI-related commercial transactions, governance processes, privacy, IP, licensing, and vendor agreements, signaling a larger enterprise rollout of AI that needs dedicated legal oversight.

Thomson Reuters says it has developed its own AI model, Thomson, and is integrating AI more deeply into CoCounsel, Westlaw, and Practical Law. The move raises the competitive bar for standalone legal AI vendors by tying AI to the company’s existing legal content and research platforms.

Thomson Reuters published a white paper arguing that legal departments have moved from AI experimentation to AI accountability. It предлагает a framework for measuring AI value through cost savings, time savings, and strategic impact, with KPIs like outside counsel spend, turnaround time, error rates, and adoption rates. The piece is more guidance than breaking news, but it reflects how corporate legal teams are now being pushed to justify AI tools with hard metrics.

Sequoia Capital is reportedly aiming to deploy about $10 billion into dedicated AI investments, its largest concentrated wager in the firm’s history. The article says the firm has already backed OpenAI at a $300 billion valuation and increased stakes in Anthropic, xAI, and legal AI startup Harvey, signaling continued capital flow into legal-tech AI.

Legora’s Built In Colorado job posting says the legal-AI company has surpassed $100M in ARR, has 1,000+ customers in 50+ countries, and has expanded through acquisitions including Qura, Walter AI and Graceview. It also names law firms and Barclays among customers, signaling continued enterprise adoption and growth.