ChatGPT for Lawyers: Drafting, Research, and Verification

ChatGPT can help lawyers draft faster, spot issues, and organize their thinking on legal problems. It produces confident-sounding text quickly, which makes the verification step more important, not less. A tool like fenbs keeps the non-sensitive working rules, approach decisions, and next steps that apply across a type of matter outside any single chat, so each drafting or research session starts from the current approved approach rather than rebuilding context from scratch.

This guide focuses on how lawyers can use ChatGPT within the professional and ethical constraints that apply to legal work. Every jurisdiction's rules are different; the principles here reflect generally applicable obligations, but you should always check your own bar's guidance on AI tools and attorney obligations.

Where ChatGPT fits in a legal workflow

The strongest uses for ChatGPT in legal practice are tasks that produce a draft or a structured starting point for an attorney to evaluate and verify. The model works well for drafting first versions of familiar document types, summarizing legal concepts, and generating issue checklists for attorneys to review.

• First drafts of standard contract clauses, NDAs, or engagement letters based on a clear prompt.
• Memo outlines and research frameworks for common legal questions.
• Plain-language summaries of complex provisions for client communications.
• Issue spotting checklists for transactions, disputes, or regulatory matters.
• Cover letter and brief section drafts that attorneys then develop with actual authority.
• Proofreading drafts for defined-term consistency, missing cross-references, or structural gaps.

The non-negotiable: verify every legal authority

ChatGPT does not access Westlaw, LexisNexis, or other legal databases. It generates text based on its training data, which has a cutoff date and does not include real-time legal developments. More critically, it sometimes generates case citations that look real but do not exist — a phenomenon called hallucination.

Several lawyers have faced bar discipline and court sanctions for submitting AI-generated briefs containing fabricated citations. These were not careless practitioners; they were experienced attorneys who trusted the confident presentation of incorrect information. Verifying every citation is not an optional extra step when using AI for legal drafting.

The verification rule is simple: if you are citing a case, find it yourself in a primary legal source, read the relevant section, and confirm the holding supports your point. If you are citing a statute or regulation, check the current text in the official code. ChatGPT can help you draft the argument; you find the actual authority.

Record verified authorities as you find them. In fenbs, save confirmed citations against the project or task, with a brief note on what each authority establishes. That library builds over a matter and is available when you return to draft.

Confidentiality and privilege

Your professional obligations regarding confidentiality and attorney-client privilege apply to AI tools just as they do to other third-party services. Pasting client names, confidential case details, privileged communications, or personally identifying information into a public AI tool raises real ethical concerns in most jurisdictions.

The ABA and several state bars have addressed how competence and confidentiality obligations apply to AI tools. Check your jurisdiction's specific guidance before using ChatGPT or any other public AI tool for client work. Most state bar ethics opinions are publicly available on bar association websites.

The practical solution is to use hypothetical or anonymized details instead of real client information. Describe the legal situation and the type of transaction or dispute without naming the client or including identifying details. For most drafting and research tasks, this works well: the legal analysis does not depend on the client's name.

Drafting contracts and transactional documents

For standard transactional documents, ChatGPT can produce a first draft that saves meaningful time. The prompt should include the document type, jurisdiction, the basic structure you need, and any specific provisions to include. The more specific you are, the closer the first draft will be to what you need.

Review the draft carefully for defined terms used before they are defined, cross-references that do not exist, provisions that contradict each other, and standard clauses that are missing for this document type in this jurisdiction. ChatGPT drafts confidently but does not have a systematic check for those errors.

For firms that draft the same document types regularly, a set of approved prompt instructions that produce a reliable first draft is more valuable than a one-time shortcut. The prompt becomes a firm standard: every attorney using it starts from the same baseline, and improvements to the prompt improve every future draft.

Client intake and matter setup

ChatGPT can help draft the documents that start a client relationship: engagement letters, client intake questionnaires, and onboarding communications. A clear engagement letter that covers the scope, the fee arrangement, the client's responsibilities, and the billing process sets expectations that prevent problems later. Ask ChatGPT for a first draft and then review it against your firm's standard terms and the specific circumstances of the matter.

For conflict check summaries, ChatGPT can help organize the information you have gathered into a clear description of the parties, the nature of the matter, and the potential issues for the responsible attorney to review. It does not have access to your firm's conflict database; the actual conflict search is a firm process. But drafting a clear summary description speeds the review conversation.

Keep approved engagement letter templates and intake question lists in a shared location. The next attorney opening a similar matter has a starting point rather than drafting from scratch. Updating the template when the firm's standard terms change is more reliable than each attorney maintaining their own version separately.

Billing descriptions and time entries

Writing billing descriptions that accurately summarize work performed is routine but time-consuming. ChatGPT can help draft billing descriptions from your notes or from a brief description of what you did. Give it the task type, the time spent, and the key actions taken. Review the draft to confirm it accurately describes what you actually did and meets your jurisdiction's billing description standards.

Never use ChatGPT to pad billing entries or overstate work performed. Bar ethics rules address billing practices directly, and inaccurate descriptions create professional responsibility exposure. Use ChatGPT to write clearly and accurately about what you actually did, not to add apparent substance to entries that do not reflect real work. Clear billing descriptions also help the client understand what they are paying for, which reduces billing disputes and strengthens the client relationship over time. A client who can follow what their attorney did is more likely to pay promptly and return for future work than one who receives entries they cannot interpret. Clear billing records also serve as a useful reference when reviewing how time was spent across the different phases of a matter.

Research memos and brief sections

ChatGPT can produce a useful research memo outline: the main legal questions, the relevant analytical framework, and a list of issues to resolve. That structure guides the actual research without substituting for it.

For brief writing, ChatGPT can help with argument structure and plain-language clarity once you have the actual authorities. Draft your argument with real citations, then use ChatGPT to improve sentence clarity, check logical flow, and suggest counterarguments you may need to address.

How to do this with fenbs

Use fenbs to track the approach and verified authorities for each matter type, keeping non-sensitive working rules accessible across sessions.

1. Create a free account and set up a project for the matter type or specific matter you are working on.
2. Record the approach rules as Decisions: jurisdiction, governing law, standard provisions, any firm preferences. Keep all confidential client details out of this record.
3. As you verify sources, add them as links with brief notes on what they establish. Build the authority library as you work through the matter.
4. Connect your preferred AI app from Integrations and ask it to read the project context before drafting or researching.
5. After each session, record what was completed, what was verified, and what still needs checking before you close the work.

fenbs does not access legal databases or give legal advice. It keeps the non-sensitive working rules and verified source links that help each session build on confirmed work. Legal judgment stays with the attorneys responsible for the matter.