Contract Drafting and Review

---

What You'll Learn

By the end of this module, you will be able to:

• Use AI to draft contract clauses tailored to specific deal requirements
• Review contracts for risks, missing provisions, and one-sided terms
• Apply prompt templates for the most common contract types
• Implement a practical redlining workflow using AI tools
• Understand the limitations of AI in contract work, including jurisdiction-specific requirements it may miss

Estimated Time: 1-2 hours

---

3.1 Using AI to Draft Contract Clauses

Why AI is Well-Suited for Contract Drafting

Contract drafting is one of the areas where AI tools deliver the most immediate value. Contracts are highly structured documents with predictable patterns — exactly the kind of content LLMs handle well. AI can:

• Generate first drafts of standard clauses in seconds
• Adapt language to different deal structures and risk profiles
• Produce multiple alternative versions of a clause for negotiation
• Ensure consistency in defined terms and cross-references throughout a document

The Right Mental Model

Think of AI as a highly efficient first-draft machine. It produces a starting point that you then review, refine, and customize. The AI draft is never the final product. Your professional judgment, knowledge of the client's specific needs, and understanding of the deal context are what transform a generic AI draft into a quality legal document.

Drafting Standard Clauses

When asking AI to draft contract provisions, specificity in your prompt produces dramatically better results.

Weak Prompt:

Write a confidentiality clause.

Strong Prompt:

Draft a mutual confidentiality clause for a technology licensing
agreement between two US-based companies. The clause should:

1. Define confidential information broadly to include technical
   specifications, source code, customer lists, and financial terms
2. Include standard exclusions (publicly known, independently
   developed, rightfully received from third parties)
3. Set the confidentiality period at 3 years after disclosure
4. Allow disclosure to employees and contractors on a need-to-know
   basis, provided they are bound by similar obligations
5. Include a carve-out for legally compelled disclosure with
   prompt notice to the disclosing party
6. Specify that the receiving party's obligations survive
   termination of the agreement

Use clear, modern drafting style. Avoid archaic legal language.
The governing law is Delaware.

The second prompt gives the AI enough context to produce a clause that is much closer to what you actually need, saving significant revision time.

---

3.2 Reviewing Contracts for Risks

AI as a Second Set of Eyes

Contract review is time-intensive and detail-oriented. AI can serve as a powerful first-pass review tool, identifying provisions that need attention before you begin your detailed analysis.

Prompt Template 1: General Contract Risk Review

I am reviewing the attached [contract type] from the perspective
of [your client's role: buyer/seller/licensee/landlord/etc.].

Please analyze this contract and identify:

1. **High-risk provisions** — clauses that create significant
   liability, broad indemnification obligations, or unfavorable
   limitation of liability terms
2. **Missing provisions** — standard protections for [client's role]
   that are absent from this agreement
3. **One-sided terms** — provisions that disproportionately favor
   the other party
4. **Ambiguous language** — clauses that could be interpreted in
   multiple ways, creating potential disputes
5. **Unusual provisions** — anything that departs from market
   standard terms for this type of agreement

For each item identified, explain the risk and suggest specific
alternative language.

How to Use This Template: Copy the contract text into your AI tool along with this prompt. For longer contracts that exceed the AI's context window, break the document into sections (e.g., review the indemnification and liability sections separately from the IP provisions).

Focused Review for Specific Issues

Sometimes you need to focus on particular aspects of a contract rather than a general review.

Prompt Template 2: Targeted Provision Analysis

Review the following [specific clause type, e.g., "indemnification
provision"] from a [contract type].

My client is the [role]. Please analyze:

1. What obligations does this provision create for my client?
2. What triggers the indemnification obligation?
3. Are there any caps or limitations on liability?
4. Is there a duty to defend, or only a duty to indemnify?
5. How does this compare to market standard terms for similar
   agreements?
6. What specific changes would you recommend to better protect
   my client's interests?

Here is the provision:
[paste the specific clause]

---

3.3 Common Contract Types AI Can Help With

Non-Disclosure Agreements (NDAs)

NDAs are among the most standardized contracts, making them ideal for AI assistance. AI can quickly generate NDAs tailored to different contexts:

• Mutual vs. one-way NDAs
• Employee NDAs with non-compete and non-solicitation components
• Due diligence NDAs with deal-specific provisions
• Technology evaluation NDAs with return-or-destroy obligations

Key Review Points AI Can Flag:

• Overly broad definitions of confidential information
• Missing or inadequate exclusions
• Unreasonable duration periods
• Lack of remedies provisions (injunctive relief)
• Missing residuals clauses

Service Agreements

AI is effective at drafting and reviewing service agreements across various contexts, including consulting agreements, SaaS terms, and professional services contracts.

Key Review Points AI Can Flag:

• Scope of services ambiguity
• Payment terms and late payment consequences
• Intellectual property ownership and work-for-hire provisions
• Service level commitments and remedies for failure
• Termination rights and transition assistance obligations

Employment Contracts

Employment agreements require particular care because they intersect with jurisdiction-specific employment law.

Key Review Points AI Can Flag:

• At-will vs. fixed-term employment provisions
• Compensation structure and bonus eligibility
• Non-compete scope, duration, and geographic limitations
• Intellectual property assignment clauses
• Change of control and severance provisions

Important Limitation: Employment law varies dramatically by jurisdiction. Non-compete enforceability, for example, ranges from fully enforceable in some states to completely prohibited in others (California, for instance, generally prohibits non-compete agreements). AI may not accurately reflect the current law in your specific jurisdiction.

---

3.4 Redlining and Markup Workflows

Using AI for Redlining

Redlining — marking up a contract to propose changes — is a core legal workflow. AI can accelerate this process significantly.

Prompt Template 3: Redline Generation

I have received the following contract from opposing counsel.
My client is the [role]. I need to propose redlines that protect
my client's interests while maintaining a reasonable negotiating
position.

For each proposed change:
1. Quote the original language
2. Provide the proposed revised language
3. Briefly explain why the change is important
4. Indicate whether this is a "must-have" change or a
   "nice-to-have" negotiating point

Focus on the most impactful changes first. Prioritize provisions
related to [list your top priorities, e.g., liability caps,
IP ownership, termination rights].

Here is the contract:
[paste contract text]

A Practical Redlining Workflow

Step 1: Initial AI Review Run the contract through AI with the risk review prompt (Template 1 above) to identify the issues that need attention.

Step 2: Generate Redline Proposals Use the redline prompt (Template 3) to generate proposed language changes for the issues identified.

Step 3: Professional Review and Refinement Review every AI-suggested change with your professional judgment:

• Does this change accurately reflect your client's priorities?
• Is the proposed language legally sound in your jurisdiction?
• Are the characterizations of risk accurate?
• Is the negotiation strategy appropriate for this deal?

Step 4: Finalize in Your Document Transfer the approved changes to your word processor, using track changes to create the formal redline.

---

3.5 Drafting Complete Contract Sections

Beyond Individual Clauses

Once you are comfortable using AI for individual clauses, you can use it to draft more complete contract sections or even full first drafts.

Prompt Template 4: Contract Section Drafting

Draft a complete [section name, e.g., "Intellectual Property Rights"]
section for a [contract type] between [party descriptions].

Context:
- [Party A] is providing [description of services/products]
- [Party B] is the [customer/licensee/etc.]
- The key IP concern is [describe the main IP issue]
- Governing law: [jurisdiction]

The section should include provisions addressing:
1. Ownership of pre-existing IP
2. Ownership of newly created IP/work product
3. License grants (scope, exclusivity, territory, duration)
4. Third-party IP and open-source components
5. IP representations and warranties
6. IP indemnification

Use clear, modern drafting conventions. Define terms consistently
with the rest of the agreement. Number paragraphs for easy
cross-referencing.

---

3.6 Limitations of AI in Contract Work

What AI May Miss

While AI is a powerful contract drafting and review tool, it has important limitations:

Jurisdiction-Specific Requirements:

• Many contract types have jurisdiction-specific requirements (e.g., specific language required for enforceability of liquidated damages clauses, consumer protection disclosures, or non-compete restrictions)
• AI may not be current on recent legislative changes affecting contract enforceability
• Local court interpretations of common contract language may differ from what AI assumes

Industry-Specific Standards:

• Certain industries have standard contract frameworks (e.g., ISDA for derivatives, AIA for construction) that AI may not follow precisely
• Regulatory requirements specific to your industry may not be reflected in AI drafts

Deal-Specific Context:

• AI does not know your client's business priorities, negotiation leverage, or relationship dynamics
• It cannot assess whether a particular risk is acceptable given the overall deal value and strategic importance
• It may suggest changes that are technically beneficial but commercially unrealistic

Formatting and Integration:

• AI-generated provisions may not integrate smoothly with your firm's standard templates
• Defined terms may conflict with existing definitions in the agreement
• Cross-references and section numbering may not align with the overall document structure

Best Practices for AI-Assisted Contract Work

1. Always review AI output against your jurisdiction's requirements. Do not assume the AI knows your local law.
2. Check defined terms for consistency. AI may use different terms than the rest of your document.
3. Verify any legal standards referenced. If the AI claims a provision is "market standard," confirm that independently.
4. Consider the commercial context. AI optimizes for legal protection; you need to balance protection with deal viability.
5. Maintain your firm's templates and precedents. Use AI to enhance your existing forms, not replace them entirely.

---

Key Takeaways

• AI excels at generating first drafts of contract clauses. Specific, detailed prompts produce dramatically better results than vague requests.
• Contract review is one of AI's strongest use cases. Use it as a first-pass tool to identify risks, missing provisions, and one-sided terms before your detailed analysis.
• AI can handle the most common contract types effectively. NDAs, service agreements, and employment contracts are all well within AI's capabilities, but jurisdiction-specific requirements always require human verification.
• Build a systematic redlining workflow. Combine AI-generated risk analysis with AI-proposed language changes, then apply your professional judgment to every suggestion.
• AI does not know your jurisdiction, your client, or your deal. It provides legally plausible output that must be filtered through your knowledge of the specific context.
• Use AI to enhance your contract practice, not replace your judgment. The most effective approach is using AI to handle the time-consuming first-draft and first-pass review work, freeing you to focus on the strategic and judgment-intensive aspects of contract work.