Read time: under 9 minutes

This week, the Otter AI ruling has put AI notetakers under fresh scrutiny, with privacy and wiretap claims testing whether vendors can be treated as more than a customer’s passive tool when meeting data is used for training. At ILTACON, eDiscovery teams were asking a related question: if AI tools generate inputs, outputs, and artifacts that may matter later, how should companies preserve them before litigation arrives?

The product race is showing no signs of slowing down. Thomson Reuters launched its own proprietary frontier model built on legal, tax, and news content, while Harvey introduced Tenet, its first in-house model for longer legal work. Conduent is embedding Gemini into Viewpoint for eDiscovery and compliance workflows, and Twin1 has emerged from stealth with encrypted “digital twins” for lawyers working across Slack, Teams, Outlook, Gmail, and SharePoint.

Our guest article this week takes the IP question head-on. Apple v OpenAI asks whether trade secret law is enough when employees move, institutional knowledge travels with them, and the most valuable assets in AI are increasingly hard to keep inside the building.

➡️ BIPA And Wiretap Claims Against Otter AI To Test AI Notetaker Vendor Risk | Legal commentary on the recent Otter AI ruling notes it may reshape how enterprises assess AI notetaker tools. The California court distinguished Otter from Graham v. Noom, finding Otter's use of meeting data to train its models made it more than a mere extension of the customer. It also allowed BIPA claims over speaker-tagging voiceprints to proceed, rejecting Otter's extraterritoriality defence. BIPA damages of $1,000 per negligent and $5,000 per intentional violation create meaningful exposure.
^{Aug 25, 2026, Source: Mondaq}

➡️ AI Preservation Emerges As Key Discovery Challenge At ILTACON 2026 | Companies are beginning to reckon with the need to preserve data from AI systems for potential litigation and investigations, panellists at ILTACON 2026's e-discovery roundtable said. Reed Smith counsel Marcin Krieger warned that avoiding Rule 37 sanctions starts with information governance policies covering AI tool data. HaystackID's new TRACE Suite is designed to collect and preserve AI inputs, outputs, and artifacts. Panellists also flagged that preservation supports authentication of evidence as synthetic media becomes harder to distinguish visually.
^{Aug 24, 2026, Source: Law.com}

➡️ Thomson Reuters Launches Proprietary Frontier Model Built On Legal And Tax Data | Thomson Reuters has unveiled Thomson, its first proprietary large language model, developed in-house with a $40 million investment on an open-source foundation. Trained on decades of content from Westlaw, Practical Law, Checkpoint, and Reuters, and shaped by hundreds of subject matter experts, early evaluations reportedly put Thomson on par with leading frontier models at a fraction of typical cost. Thomson debuts inside Tabular Analysis in CoCounsel Legal, with wider rollout planned.
^{Aug 24, 2026, Source: PR Newswire}

➡️ Conduent Embeds Google Gemini Into Viewpoint For Legal AI Workflows | Conduent has announced a collaboration with Google Cloud to embed Gemini models into its Viewpoint platform, targeting eDiscovery, compliance, and breach response workflows. The Enhanced Review capability applies user-defined protocols to identify relevant content and surface high-risk documents, delivering a claimed 30–60% reduction in document-intensive analysis effort. Viewpoint Apps convert unstructured documents into structured outputs such as chronologies and privilege logs.
^{Aug 21, 2026, Source: Futurum}

➡️ Twin1 Raises $20m Seed To Build Digital Twins For Lawyers | Twin1, led by former Eigen CEO Lewis Liu, has emerged from stealth with a $20 million Seed round co-led by Bessemer Venture Partners, alongside Tribeca Venture Partners and Aramco Ventures. The platform builds encrypted "digital twins" that learn each professional's knowledge, tone, and communication patterns, operating within Slack, Teams, Outlook, Gmail, and SharePoint. Linklaters, Orrick, and Dechert are among early users. The company emphasises sovereign AI and privacy safeguards.
^{Aug 20, 2026, Source: Artificial Lawyer}

➡️ Harvey Launches Tenet, Its First Proprietary AI Model For Legal Work | Harvey has introduced Harvey Tenet, its first in-house large language model, designed to handle multi-hour legal work at lower cost than third-party models. Trained on a version of Chinese open-source model Kimi K3, Tenet was shaped using data created by attorneys via Mercor and Snorkel. It launches within a broader "Harvey II" rollout that includes a new Memory feature. Co-founder Gabe Pereyra says Tenet could serve as a base for firm-specific models.
^{Aug 18, 2026, Source: Business Insider}

The last several months have produced a striking run of intellectual property theft allegations across the AI industry. Apple has accused OpenAI of a coordinated effort to extract its confidential information, and in a parallel debate several U.S. labs have accused Chinese developers of distilling American models to train their own. The two stories are different but have similar takeaways: the most valuable assets in technology are increasingly mobile, and they are getting harder to keep inside the building when employees leave. The Apple case offers the clearest example.

On July 10, 2026, Apple sued OpenAI, io Products and two former Apple employees, Tang Tan and Chang Liu, alleging they misappropriated product designs, manufacturing processes and supply-chain strategies. Filed under the Defend Trade Secrets Act and breach of their intellectual property agreements in the Northern District of California, the complaint states that more than 400 former Apple employees now work at OpenAI. That number is the bigger story.

In California, organisations generally cannot use non-compete agreements to stop employees from joining rivals. Employees are free to leave, and the expertise, judgment, and institutional knowledge in their heads goes with them. Trade-secret law can stop employees from taking confidential files or using protected information, but it cannot prevent them from applying their general skills and experience elsewhere. As a result, Apple has faced a reactive remedy – proving after the fact through laptops, emails, downloads and circumstantial evidence, that specific information was taken or used improperly. This type of dispute can be hard to prove and quickly evolve into an expensive argument over what was stolen, what was remembered and what was independently developed.

The nature of the assets Apple claims to have been taken is important. Schematics, product designs, manufacturing processes, materials techniques and a specific metal-finishing process may all contain patentable inventions.

Click here to continue reading the full interview

Claude can be your analyst, editor, and strategist.

But most professionals are using it to fix grammar.

These 1,000+ Claude prompts take it from grammar tool to your most powerful AI work assistant.

Sign up for Superhuman AI and get:

Claim your free prompts

The AI Regulation Tracker offers a searchable overview that gives you instant snapshots of how each country is handling AI laws.

🆕 360 Law Group - Home to 360 Business Law and 360 Law Services, we connect exceptional legal talent, technology, and modern legal models to meet the evolving needs of businesses and individuals worldwide.

🆕 Twin1 - The AI extension of you, built for teams with professional expertise and complex coordination.

🆕 LegalFab - LegalFab.ai is the central nervous system for the connected legal enterprise. Unify your firm's systems. Orchestrate Al securely. Deliver measurable transformation across the client and matter lifecycle.

Want more Legal AI Tools? Check out our
Top AI Tools for Legal Professionals

Why it helps: Turns a broad legal question into a targeted research roadmap, helping lawyers search faster, avoid irrelevant detours, and move more quickly from issue-spotting to usable advice.

The Responsible AI Collaborative is a not‑for‑profit organization working to present real‑world AI harms through its Artificial Intelligence Incident Database.

View the latest reported incidents below:

⚠️ 2026-07-26 | Anthropic and OpenAI AI Agents Reportedly Took Unsanctioned Actions on the Live Internet During UK AISI Cybersecurity Evaluations | View Incident

⚠️ 2026-06-24 | Rep. Anna Paulina Luna's Office Reportedly Published Claude Transcript Residue on Accident in House Amendment Summary | View Incident

⚠️ 2026-06-11 | 3M-Retained Expert Reportedly Submitted Largely ChatGPT-Generated Analysis Seeking to Exonerate 3M as Expert Evidence in Explosion Lawsuit | View Incident

The Legal Wire is an official media partner of:

Thank you so much for reading The Legal Wire newsletter!

If this email gets into your “Promotions” or "Spam” folder, move it to the primary folder so you do not miss out on the next Legal Wire :)

If you have any tips for us, just reply to this e-mail! We’d love any feedback or responses from our readers 😄

The Legal Wire takes all necessary precautions to ensure that the materials, information, and documents on its website, including but not limited to articles, newsletters, reports, and blogs ("Materials"), are accurate and complete.

Nevertheless, these Materials are intended solely for general informational purposes and do not constitute legal advice. They may not necessarily reflect the current laws or regulations.

The Materials should not be interpreted as legal advice on any specific matter. Furthermore, the content and interpretation of the Materials and the laws discussed within are subject to change.