CoCounsel Legal vs Learned Hand: how they compare in 2026

CoCounsel Legal exists only as a generative and agentic layer, rebuilt in 2026 on an agent framework so it plans, reasons and executes multi step legal work. Without the models there is no product, only the underlying Westlaw and Practical Law content, which the parent sells separately as its own products.

Learned Hand extracts facts, maps disputed issues, structures arguments and runs analysis against the procedural standard for the motion type. It generates a first draft order with its citations and checks every statement and citation against its source. The companion product, The Red Pen, reads filed briefs and tests each citation for existence, pinpoint and whether the holding supports the proposition. Beside the models, the product has an upload screen, a case management connection and an editor with a Word export.

The research agent identifies key legal concepts, selects Westlaw tools including KeyCite, the Key Number System and annotated statutes, runs multiple search cycles and synthesizes a structured result. Output grounds to primary authority the user can open, with inline citations, a Sources tab with direct links and highlighted excerpts, and KeyCite validity flags shown with the authority. A citation level verification feature, Deep Research Verify, is published, and prompting guardrails restrict output to content validated by people rather than the open internet. CoCounsel took part in the February 2025 Vals Legal AI Report, a third party benchmark against a measured lawyer baseline, scoring between 73.2 and 89.6 percent across four tasks with a 79.5 percent average, each task more than 10 points above the lawyer baseline. Thomson Reuters also publishes CoCounsel Bench, under which attorneys reviewed 50 complex tasks of about six lawyer hours each and judged the product's response stronger than the expert reference answer on nearly 40 percent. The 2025 study measured task accuracy rather than citation validity or hallucination rate, and Thomson Reuters did not take part in the later Vals study on citation authoritativeness.

Learned Hand's home page states that every citation links to its source, every claim traces to the record, any statement can be clicked through to its source, and every citation is checked against its source automatically. The vendor's release of 18 March 2026 adds that all outputs are hyperlinked to source material in the case file and that the platform runs multiple verification passes. The Red Pen's citation checks test other parties' citations in filed briefs. No accuracy figure, test set or evaluation is published, and no failure mode is named. The retrieval method and the legal sources behind the analysis are not described, and nothing states what the system does when it cannot support a proposition. The home page carries an FAQ entry on how Learned Hand prevents hallucinations.

Thomson Reuters states that its AI is built with human oversight at its core and publishes a Fiduciary-Grade AI framework built around transparency and accountability. The product is described as showing its plan before executing: it interprets a request, builds a plan specific to the matter, and returns work product with traceable citations the user can open and check. No threshold is published at which an agent stops and returns to a lawyer. The General Terms require a licensed professional or similar expert to review AI output before anyone relies on it, and nothing says what Thomson Reuters does once a wrong answer is reported. The August 2026 release moved the product from structured skills to a fully agentic experience.

Each stage of the published workflow ends with a human step. The user reviews the analysis, clicks any statement to see its source, asks questions to refine the analysis, and edits the first draft in the platform or exports it to Word. The vendor's release quotes its chief executive that the product will assist with preparation but that judges make the final decisions. Nothing is published on a threshold or mode at which the system acts without review, on any categorical limit on what a draft may be used for, or on what happens when an analysis is wrong. The home page FAQ includes an entry on what to do when a user disagrees with its analysis.

Thomson Reuters publishes case studies naming Justly Prudent, reporting five times return on investment and a doubling of litigation capacity, and Brinks, on reducing outside counsel spend. It also publishes a commissioned Forrester Total Economic Impact study dated 2026, with its method set out and its commissioned status disclosed on the page. Aggregate figures are published with the population stated, including 2,700 corporate customers using the CoCounsel family and a 61 percent reduction in time to draft correspondence.

The vendor's release of 18 March 2026 names the Superior Court of Los Angeles County as a partner, carries statements from its Presiding Judge and its Executive Officer, and states that the technology is used by the Michigan Supreme Court and by trial courts in ten states. The site adds an unattributed testimonial from a federal judge described as a user since 2024, and an unattributed Red Pen testimonial saying the platform takes minutes to do what would take a staff attorney several days. No figure for what changed at any named court is published, the ten states are not identified, and nothing describes how any result was measured.

Thomson Reuters' Legal AI Terms of 23 January 2026, part of the customer agreement, bar Thomson Reuters and its third party providers from training or fine tuning any generative or foundation model on prompts, user content or outputs unless agreed separately in writing. Prompts and outputs may be used to evaluate safety and effectiveness, with automated safeguards that keep sensitive and identifying information out of those samples. A published FAQ states plainly that customer content and prompts are not used to train or improve the product or the underlying models. Third party model partners, including OpenAI and Google, are contractually prohibited from training on customer data, and systemic controls turn off those parties' abuse monitoring so there is no human access to customer data. Thomson Reuters also publishes a zero retention API architecture and data residency controls. The General Terms treat customer data as confidential information and allow its disclosure under law, subpoena or court order only with prior notice where legally permitted. Attorney client privilege and work product handling are not addressed directly. Thomson Reuters does not describe segregation by matter, although the product is built around matters.

The Red Pen page states that filed briefs are not retained and not used to train, and that the court's data is never stored, logged or used to train models. Those statements cover a citation checker that reads briefs as they are docketed. The main workbench syncs a court's cases and holds them, and no published confidentiality statement covers it apart from a home page FAQ entry titled Is my case data safe?. No terms of service, customer agreement, privacy policy or security page exists on the site. Nothing is published on segregation between courts or users, on deliberative confidentiality, or on the position of any model provider.

The product is sold to licensed professionals and distributed to more than 200 law schools and 120,000 law students. Thomson Reuters frames the system as working at the level of a senior associate under professional supervision, and its General Terms for the United States put that supervision into the contract: a licensed professional or similar expert must review AI output before anyone relies on it, uses it or passes it on, and the customer alone is responsible for verifying it. The same terms say the service is provided for information and not as legal advice. The product page, the AI hub, the trust center, the legal blog and the terms name no jurisdiction limits and say nothing of a lawyer's competence duties.

The product is sold only to courts. Learned Hand states it is built for the judiciary and contrasts itself with tools designed for lawyers and private practice. Its release quotes its chief executive that judges make the final decisions, and the home page describes the product as decision support. Nothing addresses a judge's or clerk's duties of competence and supervision when relying on a draft, and no court rule or judicial ethics standard on generative AI is named. No jurisdictional limit is stated beyond a claim that the software is jurisdiction specific. The home page FAQ includes an entry titled Does Learned Hand replace me or my clerks?.

Thomson Reuters publishes a named standard it calls Fiduciary-Grade AI with a buying framework attached, describes a dedicated research organization of more than 1,000 specialists, states a figure for annual responsible AI investment, and answers a direct published question on bias: prompting guardrails restrict output to a clearly defined and limited set of human validated content rather than the open internet. No named internal owner of model governance, pre release testing results, or disclosure of uneven output across matter types, parties or populations is published. The bias answer describes a content restriction rather than testing for disparate output.

The product is built for impartiality and designed for the bench from the first line of code, Learned Hand says, and it is contrasted with tools built for advocacy. Nothing published says who inside the company is accountable for model behavior, what is tested before a release, or what has been found about uneven output across case types or parties. No responsible AI page, policy or evaluation exists on the site. Press reporting says the company tests for bias and accuracy and has not published results. The vendor itself publishes nothing on it.

SOC 2 Type II and ISO 27001 are maintained and ISO 42001 certification is referenced for the AI management system. Thomson Reuters publishes a zero retention API architecture, data residency controls, 24 hour security monitoring and threat detection, corporate identity management integration, and a corporate trust center carrying the information security framework. The Data Security Addendum that the General Terms apply, last modified 26 October 2026, commits to notice of a security breach within 72 hours, and the General Terms delete production data within 90 days after termination on written request. No current named subprocessor list is published for this product. The two named model partners are disclosed in the AI FAQ rather than in a maintained subprocessor list.

The Red Pen page states zero data retention: filed briefs are not retained, and data is never stored, logged or used to train models. For the workbench, which syncs a court's case management system and holds its cases, no retention period, deletion commitment, access control description, subprocessor list or incident practice is published. No privacy policy, security page or trust center exists on the site, which lists two pages in its own search results.

CoCounsel Legal is sold under Thomson Reuters' General Terms for the United States (version 6.0, 1 September 2026), which cap each party's liability in a calendar year at the fees payable for the affected service in the twelve months before the claim. The cap doubles for breaches of confidentiality, data protection or security, and indirect and consequential loss is excluded. Thomson Reuters defends and indemnifies the customer against third party claims that use of the service infringes intellectual property, and that indemnity sits outside the cap, except for claims arising from third party services Thomson Reuters licenses, which are capped at twice the prior twelve months' fees. The service warranty promises conformity with the documentation, with repair, modification or replacement as the remedy and termination if that fails. For AI output, the terms disclaim liability for its accuracy, legality or reliability and for decisions based on it, and the customer must have a licensed professional or similar expert review it before relying on it. The Legal AI Terms of 23 January 2026, which govern the AI features, add no liability term of their own, and neither document states an insurance position.

Nothing is published on who bears the loss when the system is wrong. The home page, The Red Pen page and the vendor's release of 18 March 2026 carry no terms of service, customer agreement, warranty, indemnity, limitation of liability or insurance statement, and the site's own search results list no other pages. Court agreements are reported in the press, and none is published.

Named connections span document management with iManage, NetDocuments and SharePoint, Microsoft 365, the parent's own HighQ collaboration platform and Legal Tracker, contract lifecycle through Icertis, plus DeepJudge, Supio and Smokeball, and a Word add in carrying a personal clause library. An MCP integration lets a lawyer reach the product from a separate AI client without switching platforms. No documentation describes what each integration moves, in which direction, or what an administrator must configure.

Integration with a court's case management system is the home page's promise: sync once and the cases are there. Documents can be uploaded or connected directly from the case management system, and The Red Pen page says every filing is checked automatically the moment it is docketed. Drafts can be exported to Word. No case management system is named, nothing states which direction data moves or what a court must configure, and no API or implementation documentation is published.

Published data residency controls keep information within required geographic boundaries. The rollout is sequenced by country across the United States, Canada, the United Kingdom and Australia, and FedRAMP In Process status is held for the underlying Westlaw and Practical Law products, with stated investment toward authorization for this one. The list of available regions, the tenancy model, what changes between tiers, and where processing happens as distinct from where data is stored are not published.

Nothing is published on where the software runs or where court data sits. No tenancy model, hosting provider, region or deployment option is stated anywhere on the site.

SOC 2 Type II and ISO 27001 are stated as maintained, and ISO 42001 is referenced for the AI management system. FedRAMP In Process status is published for Westlaw and Practical Law, with authorization stated as an investment goal for this product, and a corporate trust center publishes the information security management framework. No attestation report, coverage period, audit scope or named auditor is published, and no route to request the reports without a sales conversation is published either. Certifications are stated at the parent company level rather than scoped to this product.

No independent security attestation is published. No SOC 2, CJIS or other attestation appears on the vendor's site, which has no security or trust page, or in the vendor's release of 18 March 2026, and none is attributed to the vendor elsewhere.

Anthropic's Claude Agent SDK is named as the architecture of the current release, and a published FAQ names OpenAI and Google as third party AI partners and states the contractual position imposed on them. Which model serves which task, where the models run, a maintained subprocessor list for this product, and any commitment to notify customers before the model supply chain changes are not published.

Nothing is published by the vendor about the models underneath. The site contrasts the product with generic AI and with ChatGPT but names no model, provider or architecture, and a home page FAQ entry asks how Learned Hand is different from ChatGPT. Press reporting attributes to the company the use of models from Anthropic, OpenAI and Google, which is not the vendor's own published material.

Tier names and feature splits are published and reachable from the main navigation: CoCounsel Essentials, Westlaw Advantage with CoCounsel Essentials, Practical Law Dynamic Tool Set with CoCounsel Essentials, and CoCounsel Legal. Each has its own plans and pricing page, with variants for law firms, corporations and government. An online configurator on the Thomson Reuters sales site prices new customers at law firms of up to ten attorneys by number of attorneys, jurisdiction and plan length. For one attorney it shows CoCounsel Essentials at $360.00 a month on a one year plan and $292.50 a month on a three year plan, with 12% off for two years and 18% off for three. Firms with more than ten attorneys and existing customers are routed to a sales quote, so no figure is published for the larger part of the range.

No pricing information is published at any level. The site's own search results list no pricing page, and the home page, The Red Pen page and the vendor's release of 18 March 2026 give no price, tier, unit or term. The only route offered is a demo request. A public sector reseller lists the product on the NASPO ValuePoint and SEWP V contract vehicles without a price, and press reports give court contract values. Neither is the vendor's own publication.

Published navigation separates law firms under ten attorneys from firms over ten, and separately addresses corporations, government (including courts, prosecutors, public defenders, city attorneys and law enforcement), and academic use across more than 200 law schools. Practice areas are listed with their own pages, spanning business and commercial, civil litigation, criminal defense, family, labor and employment, mergers and acquisitions and personal injury. The published table of plans shows which content each plan reaches.

Learned Hand says it serves the judiciary only, naming judges, clerks, staff attorneys and court staff, and positions itself against tools built for lawyers and private practice. The work is described by motion rather than practice area: summary judgment and postconviction relief on the home page, pro se filings on The Red Pen page, and analysis that follows the procedural standard for the motion and the court's local rules. Its release names the Michigan Supreme Court and says trial courts in ten states use it. The ten states are not named, the supported motion types are not listed in full, and nothing says which courts, case types or jurisdictions the product does not handle.