AI in Law and Professional Services: Harvey to Big Four — 2026-09-13
Harvey AI closed a $550 million funding round at a $15.6 billion valuation, launching "Tenet," its first proprietary legal AI model designed to keep client data within firm-controlled environments. Meanwhile, courts have sanctioned attorneys for filing briefs with hundreds of AI-hallucinated citations, signaling a shift from questioning AI usage to demanding rigorous verification of outputs.
AI in Law and Professional Services: Harvey to Big Four — 2026-09-13
Top developments
Harvey Raises $550M for Proprietary Legal Model "Tenet"
On September 9, 2026, legal AI startup Harvey announced the close of a $550 million funding round, bringing its valuation to $15.6 billion. The company introduced "Tenet," its first proprietary large language model specifically engineered to ensure law firm documents remain off third-party servers, addressing critical attorney-client privilege concerns. This move follows Harvey's acquisition of Guardrails AI for agent security and solidifies its position serving 80% of the Am Law 100 firms.
Sixth Circuit Sanctions Attorneys for 24+ Fake Citations
In a significant enforcement action highlighted on September 12, 2026, reports detailed the Sixth Circuit’s March 2026 sanction of two Tennessee attorneys in Whiting v. City of Athens. The panel found counsel had filed briefs containing more than 24 fake citations generated by AI. The court explicitly rejected claims that errors were typographical, establishing a precedent that judges are now scrutinizing the verification of AI-assisted work rather than just the usage itself.
AI Hallucinations Surpass 2,000 Court Cases
Data released on September 8, 2026, indicates that generative AI hallucinations have been listed in more than 2,000 court cases. This statistic emerged as Google pitched its new legal model, highlighting the persistent risk of fabricated case law despite advances in retrieval-augmented generation (RAG). The volume of cases underscores the systemic nature of the problem, where polished but erroneous outputs continue to infiltrate litigation filings.
Harvey and Legora Integrate OpenAI’s GPT-6 Astra
On September 7, 2026, Artificial Lawyer reported that both Harvey and rival Legora have provided insights into their integration of OpenAI’s newly released GPT-6 Astra model. Described as OpenAI’s most capable LLM yet for end-to-end tasks, GPT-6 Astra is being leveraged to enhance the accuracy and reasoning capabilities of legal AI platforms. This integration marks a rapid adoption cycle for flagship models within the specialized legal tech sector.
Local view
German media continues to focus on the regulatory and operational implications of AI in law firms. On September 10, 2026, Haufe reported on the phenomenon of "Dr. KI," where clients arrive at tax and law firms with AI-generated answers. Diana Schmid and Jasmin Müller, partners at SteuerDigital in Tübingen, noted that while clients are better informed, this "AI half-knowledge" complicates advisory work by introducing inaccuracies that professionals must debunk.
Additionally, Pexon Consulting published guidance on September 10 regarding § 43e BRAO (German Federal Lawyers' Act), emphasizing that standard AVV (Data Processing Agreements) are insufficient for AI compliance. They argue for explicit consent, human-in-the-loop (HITL) protocols, and text-form commitments to protect client confidentiality under German law.
Context & numbers
The legal AI market is experiencing a valuation surge alongside rising sanctions. Harvey’s new $15.6 billion valuation contrasts with its earlier $11 billion mark, reflecting intense investor confidence despite operational risks. Rival Legora is reported to be negotiating a new round that could push its valuation above $10 billion, with estimated ARR reaching approximately $150 million by Q2 2026.
On the liability front, the cumulative financial penalties for AI-related misconduct are growing. Recent trackers indicate combined sanctions, fines, and opposing fees exceeding $109,700 in specific high-profile cases, representing some of the largest aggregate penalties recorded to date.
On the radar
- Billing Model Shifts: German legal publication beck-aktuell continues to highlight the tension between traditional hourly billing ("Stundenabrechnung") and value-based pricing as AI automates routine tasks like research and contract review. Firms are struggling to justify hourly rates when AI completes tasks in minutes that previously took hours.
- Big Four AI Hiring: Reports from mid-2026 indicate that Deloitte, PwC, EY, and KPMG are hiring AI engineers at a faster rate than traditional auditors, signaling a structural shift in professional services staffing models.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
Powered by
