Trending Now

Which AI Model Is Best for Legal Work? What 2026 Research Says About Accuracy

Which AI Model Is Best for Legal Work? What 2026 Research Says About Accuracy

Law firms, funders and legal departments are being sold AI for contract review, legal research and citation checking, and the models change every few months. So we looked only at independent studies published in 2026 that tested the current generation of models from OpenAI, Anthropic and Google on real legal tasks. The short version: the best models are now genuinely good at reading and extracting from documents you give them, still unreliable at recalling law from memory, and the commercial legal research tools lag behind the best custom systems.

At a Glance

Best overall model for legal document work: Google's Gemini 3.1 Pro. It was at or near the top in every 2026 study that tested it, and it was usually the fastest and cheapest of the leaders. OpenAI's GPT-5.5 found slightly more errors in contract review, and Anthropic's Claude models were the most careful about not flagging problems that weren't there.

Best accuracy recorded on a full legal task: 92%, on a 50-state statutory research test run by Stanford, achieved by a purpose-built research tool. The lesson is that how the AI is set up matters as much as which model sits underneath it.

Range of accuracy: from under 7% (asking a model to recall exact case citations from memory) to 99–100% (catching a citation to the wrong case when the model can read the source). Most real-world document tasks landed between 60% and 85%.

Westlaw and Lexis AI: 58% and 64% accuracy on a Stanford statutory survey test, below a custom-built tool at 83–92%.

Biggest single improvement: giving the model the actual documents instead of asking it from memory cut fabricated citations from roughly 15–40% to about 4–15%, and to under 0.2% with a well-built retrieval system.

Key Takeaways

The model to use. For contract review and extraction, start with Gemini 3.1 Pro. It matched the top performer on catching contract errors (74% vs. 75%) at about one-seventh of the cost and in 90 seconds instead of nine minutes. If catching every possible issue matters more than time or cost, GPT-5.5 with reasoning turned on found the most. For checking citations in a brief, the best 2026 results came from GPT-5 running as an agent and from Claude Code with Claude Opus, which was the most precise.

How to Use It

  • Give it the documents. Never ask a model to supply case law or citations from memory.
  • Turn on the model's "reasoning" or "thinking" mode for review work. It improved error-catching by 9 to 11 points in contract review.
  • Use it as a first pass and a second reviewer, not the final reviewer.
  • For research, use a tool that pulls from a full, current database of the law, because weak retrieval, not the model, causes many of the errors.

What to Expect

  • Contract extraction (pulling out dates, parties, termination and liability terms): about 80–84% accuracy for the best models.
  • Final contract proofreading (defined terms, cross-references, inconsistent language): the best models catch about three-quarters of errors. On a 60-page agreement, expect it to miss some.
  • Citation checking: nearly all citations to the wrong case get caught, but wrong pinpoint pages slip through 20% to 60% of the time.
  • Research answers grounded in the right documents: roughly 6–11% of answers still contain an unsupported statement.

What to Look Out For

  • Citations from memory. When asked to recall exact citations without sources, even the best model scored under 7 out of 100, and 20 of 21 models gave confident, wrong answers more than 94% of the time.
  • Right case, wrong page. Models tend to approve a citation because the case is on the right topic, even when the cited page doesn't support the point.
  • Questions with a false premise. If your question assumes something that isn't true, models often go along with it.
  • Legal research tools' marketing. Westlaw AI and Lexis+ AI trailed a custom-built tool by 19 to 25 points on a Stanford test.
  • Studies funded by vendors. Some of the best-looking results come from companies selling legal AI. Check who ran the test.

Best Practices

  • Ground every task in source documents, and require the model to quote the passage it relied on.
  • Check every citation yourself at the pinpoint page before filing. Automated checkers help but don't replace this.
  • Turn on reasoning mode for review tasks and accept that it's slower.
  • Test a model on a few of your own documents before rolling it out. Rankings change by task.
  • Re-test when a new model version arrives; this field moves in months, not years.
  • Keep a human reviewer accountable for anything that goes to a court, a client or a counterparty.

What the Studies Found

Contract proofreading. In August 2026, researchers had experienced lawyers plant errors in contracts (misused defined terms, wrong cross-references, wrong party names, contradictions) and tested ten current models on catching them. GPT-5.5 caught 75% of errors, Gemini 3.1 Pro 74%, Claude Sonnet 4.6 69% and Claude Opus 4.7 62%. GPT-5.5 cost $1.38 per contract and took about nine minutes; Gemini 3.1 Pro cost $0.19 and took about 90 seconds. Turning on reasoning mode added 9 to 11 points. Every model was far cheaper than a lawyer, and none was close to perfect.

Contract extraction. A May 2026 study tested models on pulling 26 standard fields out of contracts. Among the major models, Gemini 3.1 Pro scored highest (82%), with Claude Opus 4.6 (82%) and Claude Sonnet 4.6 (80%) close behind and GPT-5.4 at 78%. A smaller legal-specific model built by the study's authors scored 84% at far lower cost. The authors work for Onit, which makes that model.

Made-up citations and facts. A January 2026 study had expert reviewers check 2,700 legal answers from 12 models. Asked without source documents, the best models (GPT-5.2 and Gemini 3.0 Pro) cited something false about 15–17% of the time, and the worst over 30%. Giving the models the relevant documents cut that to about 4–15%. A more carefully built retrieval system brought it below 0.2% for every model.

Research with sources. A March 2026 study found that when models answer from retrieved legal texts, Gemini 3.1 Pro produced unsupported statements 5.7% of the time versus 11.3% for GPT-5.2, and that the quality of the search step mattered more than the choice of model. Its authors sell the search component that performed best. An August 2026 study of eight research setups found unsupported answers ranging from under 10% for the best to nearly half for the worst, with the worst results on questions built on a false assumption.

Westlaw and Lexis. In a February 2026 Stanford study, researchers tested legal AI tools against a Department of Labor survey of state unemployment insurance laws. Westlaw AI scored 58% and Lexis+ AI 64%, while a custom statutory research tool scored 83%, rising to 92% after the researchers found that some of its "errors" were gaps in the government's own survey.

Citation checking. A June 2026 study found more than 1,000 court filings containing fabricated citations, a number growing every year, and tested AI checkers on catching them. GPT-5, working as an agent that looks up cases, caught 83% of planted errors; Claude Code running Claude Opus 4.8 was the most precise and scored best overall. No model reliably caught wrong pinpoint cites, partly because page numbers often sit behind Westlaw and Lexis paywalls. A separate August 2026 study found models catch 93–100% of citations to the wrong case but miss many citations to the wrong page, and even GPT-5.4 with full reasoning missed 40% of wrong pinpoints in court opinions.

Citations from memory. A May 2026 study built from 1,000 real U.S. judicial opinions asked 21 models to recall exact case citations without any sources. The best, Claude Sonnet 4.5, scored under 7 out of 100.

The Bottom Line

The 2026 research is consistent: today's best models, led by Gemini 3.1 Pro, GPT-5.5 and Claude, are useful and cheap for first-pass contract review and extraction when they work from the documents in front of them. They still invent law when asked from memory and still miss wrong pinpoint citations, so a lawyer has to verify anything that leaves the building.

Sources (All 2026)

  • Bang et al., "ContractScrub: A benchmark for final review of legal contracts" (Aug. 2026), arXiv:2608.20204
  • Lincoln et al., "A Few Good Clauses: Comparing LLMs vs Domain-Trained Small Language Models on Structured Contract Extraction" (May 2026), arXiv:2605.05532
  • Dantart, "Reliability by design: quantifying and eliminating fabrication risk in LLMs" (Jan. 2026), arXiv:2601.15476
  • Butler and Butler, "Legal RAG Bench: an end-to-end benchmark for legal RAG" (Mar. 2026), arXiv:2603.01710
  • Das et al., "How Much Do Legal RAG Systems Still Hallucinate?" (Aug. 2026), arXiv:2608.14210
  • Afane et al., "Benchmarking Legal RAG: The Promise and Limits of AI Statutory Surveys" (Feb. 2026), arXiv:2603.03300
  • Liu, Stammbach and Henderson, "Who Checks the Citations? Benchmarking Legal Hallucination Detection" (June 2026), arXiv:2606.21155
  • Verma, "Is this Citation on Point?" (Aug. 2026), arXiv:2608.12571
  • Chen et al., "LegalCiteBench: Evaluating Citation Reliability in Legal Language Models" (May 2026), arXiv:2605.10186

Commercial

View All

Triple-I Launches Consumer Campaign Against Litigation Funding in Texas and Kentucky

The Insurance Information Institute has opened a new front in the fight over third-party litigation funding, launching a consumer-awareness campaign in Texas and Kentucky ahead of both states' 2027 legislative sessions. The push adds two more targets to a state-level reform wave in which more than 20 states have already enacted funding disclosure laws.

As reported by Insurance Business, Triple-I is framing the campaign around the transmission of litigation costs into insurance premiums. Michel Léonard, the organisation's chief economist, put the argument bluntly: "Litigation is outgrowing the economy. That gap doesn't stay in the courtroom. It lands in claims, reserves and premiums." Critics of such measures counter that disclosure mandates and funding restrictions limit access to justice for claimants who cannot otherwise afford to litigate.

Casualty pricing is the evidence Triple-I points to. While commercial property rates declined over the past year, casualty lines have stayed elevated, with umbrella and excess liability recording the sharpest increases and commercial auto under sustained pressure.

The underlying cost data comes from several Triple-I studies, including work with the U.S. Chamber Institute for Legal Reform and the Casualty Actuarial Society. U.S. tort costs reached $579.7bn in 2023, nearly double the $299.7bn recorded in 2014. Other liability accounted for $249.3bn and personal auto liability for $188.1bn. Triple-I estimates that legal system abuse added between $231.6bn and $281.2bn to liability losses across the decade to 2023.

Proposed DIFC Arbitration Law Reforms Would Impose Express Third-Party Funding Disclosure

Dubai's financial free zone is moving toward a statutory disclosure obligation for litigation funders. The proposed 2026 reforms to the DIFC Arbitration Law—the most substantial overhaul since the statute was enacted in 2008—would introduce express disclosure requirements for third-party funding arrangements in arbitrations seated in the Dubai International Financial Centre.

As reported by SCC Online, the reforms were examined during Dubai Arbitration Week 2026, in a session on the proposed changes to the DIFC Arbitration Law and the new DIFC Mediation Law. Justice Michael Black of the DIFC Courts framed the funding provision as a measure to strengthen transparency and accountability in arbitral proceedings. The panel, moderated by Graham Lovett of the DIAC Arbitration Court, also included DIFC Courts Chief Justice Wayne Martin, DIAC Registrar Robert Stephen, and counsel from Akin.

The funding disclosure requirement sits within a broader package. Other proposed changes would cut the window for set-aside challenges from three months to 30 days on a non-extendable basis, permit summary determination of unmeritorious claims, allow peremptory orders backed by sanctions, give statutory recognition to emergency arbitrator decisions, and extend confidentiality to the existence of proceedings. A standalone Mediation Law would make mediated settlements binding and court-enforceable.

The reforms follow a DIFC consultation that ran from 11 June to 10 July 2026. Arbitration volume at the DIFC Courts underscores the stakes: the courts handled 54 arbitration matters in 2025 and 45 by early November 2026, roughly one-third of the Court of First Instance workload.

Federal Court Removes Hellyer Metals Liquidator, Citing Failure to Pursue Litigation Funding

The Federal Court of Australia has removed the liquidator of Hellyer Metals Pty Ltd, finding a lack of independence and criticising his handling of litigation funding offers that could have financed an investigation into a disputed asset transfer.

As reported by ABC News, Justice Shaun McElwaine delivered orders and reasons on 10 November 2026 in Goshawk Mining Pty Ltd v Hellyer Metals Pty Ltd, in the matter of Hellyer Metals Pty Ltd (in liq) [2026] FCA 1676, removing Liam Bailey of O'Brien Palmer and appointing Adam Johnston in his place, with all books and records to be handed over. The judgment records that the parties behind the share transfer under scrutiny had appointed Bailey, who then retained as his solicitor the lawyer who had introduced them. He also provided a draft statutory liquidation report to directors the court described as intimately involved in the transaction, conduct the judge called "a serious error of judgement and demonstrative of a lack of independence."

The funding findings were central. Time records showed 30.35 hours of investigative work between January and May 2026, of which just 12 minutes were spent on litigation funding. From March 2026, two companies offered to discuss funding an investigation, but the court found he failed to engage meaningfully. A funding agreement was signed only after the removal proceeding had commenced, following an approach from lawyers acting for two bankrupt former directors.

Hellyer Metals entered voluntary administration in January 2026, with Bailey appointed liquidator on 22 January. Assets had been transferred to associated company Celestial Group around the time the Tasmanian government ordered the Hellyer Gold Mine temporarily shut over unpaid debts. Goshawk Mining and Saxby Capital, which obtained a Supreme Court order freezing $36.4 million, are pursuing damages exceeding $36 million. The court described the transaction as prima facie highly questionable and said removal was indeed required.