Harvey

What is Harvey?

Harvey is an artificial intelligence platform designed specifically for the legal sector. Unlike generalist tools like ChatGPT, Harvey was trained on specialized legal data and developed for tasks that lawyers and legal teams perform daily.

Founded in 2022, Harvey quickly became a reference in the Legal Tech market by being adopted by some of the largest law firms in the world, including Allen & Overy, PwC Legal, and Macfarlanes.

How it works

Harvey is built on large language models (such as GPT-4) with specialized fine-tuning in law. This means it understands legal terminology, legal document structures, jurisdictions, and forensic argumentation logic.

Typical usage flow:

1. The lawyer uploads a document (contract, pleading, law)
2. Asks questions in natural language about the content
3. Harvey analyzes, summarizes, identifies risks, and suggests alternative wording

Key features

• Contract Review: Identifies problematic clauses, omissions, and inconsistencies.
• Assisted Legal Writing: Generation of draft contracts, pleadings, and opinions.
• Legal Research: Summary of case law, legislation, and doctrine.
• Accelerated Due Diligence: Analysis of large volumes of documents in M&A operations.
• Contract Risk Analysis: Classification of clauses by risk level.
• Legal Translation: Adaptation of documents between jurisdictions and languages.
• Firm Customization: The model can be adjusted with the firm's own standards and precedents.

Who it is for

• Medium and large law firms managing large volumes of documents.
• In-house legal departments needing agility without increasing headcount.
• Contract specialty lawyers (M&A, Corporate, Real Estate).
• Compliance and regulatory teams monitoring legislative changes.
• Employment lawyers analyzing labor contracts at scale.

Real use cases

1. M&A Due Diligence: Analysis of hundreds of contracts in days, not weeks.
2. NDAs Scale Review: Check standardization and risks in confidentiality agreements.
3. Vendor Contracts Analysis: Identify abusive or out-of-standard clauses for the company.
4. Fast Precedent Search: "What are the STJ decisions on contract termination for default in 2023?"
5. Opinion Drafting: Structure the basis of the opinion with relevant legal grounds.

Differentials compared to ChatGPT for legal use

Security and privacy

Harvey was designed with the legal sector's confidentiality requirements in mind:

• Client data is not used to train the model.
• Compliance with data protection regulations (GDPR, CCPA).
• Enterprise contracts with SLAs and confidentiality terms.
• Deployment options in private environments (dedicated cloud).

Pricing

Harvey operates on a prices on request enterprise model, negotiated based on:

• Number of users
• Volume of processed documents
• Contracted modules
• Need for model customization

There is no self-service plan or public free trial.

Pros and cons

Pros:

• Real specialization in law, not just a generic chatbot.
• Adopted by top law firms as proof of quality.
• High reliability in English contract analysis.
• Integrations with DMS systems already used in law firms.

Cons:

• Coverage of Brazilian/Latin law still limited.
• No affordable plan for solo lawyers or small law firms.
• Requires human validation — does not replace the lawyer's judgment.
• Closed model, no transparency about training data.

Alternatives to Harvey

Clio Duo, Lexis+ AI (LexisNexis), Thomson Reuters CoCounsel, and ContractPodAi are competitors in the legal AI space.