AI Prompts to Review a US Non-Disclosure Agreement Clause by Clause

Using targeted AI prompts for NDA US contract reviews allows legal teams, founders, and contract managers to audit non-disclosure agreements clause by clause in minutes. By prompting large language models with specific US legal standards—such as the Defend Trade Secrets Act (DTSA) and state-level enforceability rules—you can instantly flag non-standard obligations, aggressive indemnities, and overly broad definitions. This guide provides actionable, copy-and-paste prompts designed to surface risky language and streamline your preliminary contract review workflow.

Non-Disclosure Agreements (NDAs) are often viewed as standard, boilerplate legal instruments. However, subtle variations in drafting can create significant liability, restrict future hiring, or expose your company to unfair legal claims under US federal and state laws. Utilizing generative AI for a clause-by-clause analysis ensures that no hidden landmines slip through during high-volume contract reviews.

Disclaimer: AI-powered contract analysis serves as a preliminary triage tool and force multiplier for legal operations. It does not constitute legal advice and should always be paired with review by a licensed attorney qualified in the relevant US jurisdiction.

Why Master AI Prompts for US NDA Reviews?

Modern AI models like Claude, ChatGPT, and Gemini possess vast training data on commercial contract drafting and statutory frameworks across all 50 US states. However, asking an AI a generic question like “Is this NDA good?” yields vague, unusable feedback. To get actionable legal intelligence, you must use structured AI prompts for NDA US reviews that evaluate specific clauses against recognized industry standards, such as those recommended by the Association of Corporate Counsel (ACC) or the National Venture Capital Association (NVCA).

Precision prompting helps you:

• Identify One-Sided Terms: Instantly detect unilateral obligations in agreements presented as mutual.
• Verify Statutory Compliance: Ensure statutory carve-outs—like the mandatory whistleblower immunity under the Defend Trade Secrets Act (DTSA)—are explicitly included.
• Prevent Sneaky Clauses: Uncover hidden non-solicitation, non-compete, or injunctive remedies tucked deep inside basic confidentiality definitions.
• Draft Revisions Faster: Instruct the AI to generate specific fallback redline language tailored to US commercial norms.

Setting Up Your AI System Prompt (The Foundation)

Before running individual clause reviews, establish the system context for the AI. This primes the model to adopt the persona of an experienced US commercial transactions attorney representing your specific position (Disclosing Party, Receiving Party, or Mutual).

Copy and paste this foundational system prompt at the start of your AI session:

You are an expert US corporate and commercial transactions attorney. Your task is to analyze a Non-Disclosure Agreement (NDA) clause by clause. 

Role Parameters:
1. Jurisdiction: Focus primarily on US federal law (e.g., Defend Trade Secrets Act) and state contract law standards (e.g., Delaware, New York, California).
2. Position: Represent the [INSERT: Receiving Party / Disclosing Party / Mutual Party].
3. Analysis Objective: Identify non-standard terms, unreasonable risks, legal ambiguities, and missing standard carve-outs.
4. Output Style: Provide a risk rating (Low, Medium, High), an explanation of the commercial risk, and exact redline replacement text for any high or medium-risk language.

Do not perform general summaries. Analyze only the text provided in each prompt according to these instructions. Confirm you understand these instructions.

Clause-by-Clause AI Prompts for US Non-Disclosure Agreements

1. Definition of Confidential Information

The definition of Confidential Information dictates the entire scope of the agreement. Disclosers want an expansive definition, while Receivers need strict parameters, such as marking requirements or clear category descriptions, to avoid accidental breaches.

Act as a US commercial legal expert. Analyze the following "Definition of Confidential Information" clause for a US-based commercial NDA.

Party Represented: [Receiving Party / Disclosing Party]

Evaluate the clause against the following rules:
1. Does it require written materials to be marked as "Confidential" or "Proprietary"?
2. Does it require oral disclosures to be summarized in writing within 30 days?
3. Is the definition overly broad (e.g., covering general business knowledge or publicly accessible data)?
4. Does it improperly include trade secrets without separating them from general confidential data?

Draft Clause:
"[INSERT CLAUSE TEXT HERE]"

Output Format:
- Risk Level: [Low / Medium / High]
- Risk Analysis: [2-3 sentences]
- Suggested Revision: [Provide balanced replacement text]

2. Standard Exclusions and Carve-Outs

Under US contract law, a valid NDA must exclude information that the receiving party already knows, rightfully acquires elsewhere, or independently develops. Missing any of the standard exclusions creates unacceptable operational liability.

Review the following "Exclusions from Confidential Information" clause under US contract law standards.

Check if ALL five standard US exclusions are present and clearly drafted:
1. Information already in the public domain through no fault of the Receiving Party.
2. Information already in the Receiving Party's possession prior to disclosure without duty of confidentiality.
3. Information rightfully disclosed to the Receiving Party by a third party without restriction.
4. Information independently developed by the Receiving Party without reference to or reliance on the Disclosing Party's Confidential Information.
5. Information required to be disclosed by law, legal process, or court order (with prompt notice requirements).

Clause to Analyze:
"[INSERT CLAUSE TEXT HERE]"

If any exclusion is missing or overly restricted (e.g., requiring "clear and convincing evidence" instead of standard "written records"), flag it as a High Risk and provide redline text.

3. Standard of Care and Duty of Secrecy

US commercial contracts generally enforce a standard of “reasonable care” or the same degree of care the party uses for its own confidential information. Demanding strict liability or standardless protection creates significant risk for the receiving party.

Analyze the following "Standard of Care" clause in a US non-disclosure agreement.

Check for the following risks:
1. Does the clause impose a strict liability or absolute duty of security rather than a "reasonable degree of care" standard?
2. Does it require compliance with specific security frameworks (e.g., SOC 2, ISO 27001, NIST) that may not align with standard business operations?
3. Is there an obligation to notify the Disclosing Party of a suspected breach within an unreasonably short timeframe (e.g., "immediately" vs. "within 5 business days")?

Draft Clause:
"[INSERT CLAUSE TEXT HERE]"

Provide an assessment and redline the language to reflect standard US commercial reasonableness.

4. Term, Duration, and Trade Secret Survival

Under US law, trade secrets enjoy perpetual protection so long as they remain secret, whereas standard commercial information usually carries a finite confidentiality term (typically 1 to 3 years). Mixing these two concepts can invalidate protections or trap receiving parties indefinitely.

Evaluate this "Term and Termination" clause for a US NDA.

Primary Checks:
1. What is the explicit duration of the confidentiality obligation (e.g., 1 year, 2 years, 5 years, perpetual)?
2. Is there a clear distinction between standard Confidential Information and Trade Secrets? (Trade secrets should be protected for as long as they qualify as trade secrets under applicable state/federal law).
3. Does the agreement contain an indefinite confidentiality term for general business data? (Flag this as High Risk for the Receiving Party).

Clause to Review:
"[INSERT CLAUSE TEXT HERE]"

Suggest redline language that aligns with Delaware/New York commercial practice.

5. Statutory Carve-Outs (DTSA Whistleblower Protections)

The US Defend Trade Secrets Act (18 U.S.C. § 1833(b)) mandates that contracts containing confidentiality obligations must provide explicit notice of immunity for whistleblower disclosures to legal authorities. Failure to include this notice forfeits the right to recover exemplary damages and attorney fees under the Act.

Inspect the following NDA text to verify compliance with US federal whistleblower laws.

Check Requirements:
1. Does the agreement include express language protecting the receiving party/employees for disclosing trade secrets to government officials or an attorney solely for reporting or investigating a suspected violation of law?
2. Is there explicit reference to 18 U.S.C. § 1833(b) or equivalent DTSA whistleblower immunity language?

Clause Text:
"[INSERT CLAUSE TEXT HERE]"

If compliant, reply: "COMPLIANT". If non-compliant, explain the statutory risk under the DTSA and provide the precise statutory immunity clause to insert.

6. Governing Law, Jurisdiction, and Dispute Resolution

Choice of law and jurisdiction determine where and how lawsuits are fought. Selecting an unusual state or agreeing to unfavorable dispute procedures can dramatically increase legal costs.

Review the "Governing Law and Jurisdiction" clause below.

Assessment Criteria:
1. Is the chosen state law standard for US commercial transactions (e.g., Delaware, New York, California, Texas)?
2. Does it mandate exclusive jurisdiction in a specific state/court?
3. Does the clause include a waiver of jury trial?
4. Does it mandate arbitration? If so, does it specify established rules (e.g., AAA or JAMS)?

Clause:
"[INSERT CLAUSE TEXT HERE]"

Highlight any procedural risks or unfavorable jurisdictions for a party located in [INSERT YOUR STATE].

7. Non-Solicitation and Restrictive Covenants (Hidden Risks)

Unscrupulous drafters often embed non-solicitation, non-hire, or non-compete clauses inside routine NDAs. In many US states—such as California—non-compete clauses are unenforceable under statutes like California Business and Professions Code § 16600, and restrictive non-solicits may be challenged.

Audit the NDA excerpt below for hidden non-solicitation, non-compete, or non-circumvention provisions.

Analyze:
1. Is there an obligation preventing employee solicitation or hiring?
2. Is there a non-circumvention clause restricting direct business deals with third-party contacts?
3. Are these restrictions reasonable in scope, geographic reach, and duration (typically 1 year maximum)?
4. Is this NDA governed by California, Minnesota, or Oklahoma law where such covenants are severely restricted or void?

Text Segment:
"[INSERT CLAUSE TEXT HERE]"

If restrictive covenants are found, classify as HIGH RISK for standard introductory NDAs and provide fallback language to delete or narrow the restriction.

Comparing Standard vs. High-Risk US NDA Provisions

When running ai prompts for nda us reviews, compare your output against these standard benchmark terms commonly used in US commercial transactions:

Best Practices for Using AI in US Legal Contract Audits

To maximize accuracy and maintain regulatory compliance when using AI prompts for NDA reviews in the United States, adopt these operational guidelines:

• Protect Data Privacy: Never upload raw non-disclosure agreements containing sensitive personal identifiable information (PII), proprietary source code, or material non-public information (MNPI) into public consumer AI models. Use enterprise-grade, zero-data-retention AI tools or local offline models.
• Contextualize State Law Specifics: US contract law is state-specific. If your contract selects Delaware, New York, or California, explicitly include that state in your prompt context, as legal definitions of trade secrets and restrictive covenants vary significantly across state lines.
• Enforce Redlines: Always instruct the AI to output exact redline text using standard legal phrasing (e.g., “Delete section X and insert section Y”). This makes applying edits in Word or Google Docs effortless.
• Cross-Check Statutory Citations: While LLMs excel at text comparison, they can occasionally hallucinate code section numbers or case citations. Always verify statutory references with legal databases before finalizing documents.

Common Pitfalls when Reviewing NDAs with Generative AI

Even advanced AI systems make structural mistakes during legal workflows. Watch out for these three frequent errors:

1. Treating Silent Clauses as Safe: AI models analyze the text provided. If an NDA completely omits crucial standard exclusions (like independent development), an AI might pass the clause as low-risk unless your prompt specifically instructs it to look for missing terms.
2. Conflating Trade Secrets with Confidential Information: General business plans lose confidentiality protection once the contract term ends, whereas trade secrets maintain legal protection indefinitely under the Uniform Trade Secrets Act (UTSA) and DTSA as long as secrecy is maintained. Prompts must force the AI to keep these classifications separate.
3. Over-reliance on Automated Risk Scores: A clause flagged as “Low Risk” by AI may still contradict internal company policy (such as insurance requirements or specific security standards). Always review outputs against your corporate playbook.

Frequently Asked Questions

Can AI replace a US licensed attorney for NDA reviews?

No. Generative AI is an efficiency tool for contract triage, initial redlining, and risk spotting. It cannot provide formal legal advice, account for nuanced litigation risks, or execute legal strategy. Critical agreements should always be finalized by a qualified US attorney.

What specific US federal laws should be checked in an NDA?

The primary federal statute to verify is the Defend Trade Secrets Act (DTSA) of 2016 (18 U.S.C. § 1836 et seq.), specifically the mandatory whistleblower immunity provisions under 18 U.S.C. § 1833(b). Federal securities laws regarding insider trading should also be considered if evaluating publicly traded companies.

Is it safe to upload confidential agreements into AI tools?

Public consumer platforms may use uploaded content to train future models. For confidential contract reviews, companies must utilize enterprise AI subscriptions with strict zero-data-retention agreements, HIPAA/SOC 2 compliance, and opt-outs from model training.

How do I prompt AI to flag non-compete clauses banned by state laws?

Use targeted prompts that instruct the model to review the agreement under specific state laws, such as California Business and Professions Code Section 16600 or similar rules in Minnesota and Oklahoma. Direct the AI to flag any non-competition or restrictive employee covenants as void or illegal under those jurisdictions.

Conclusion

Employing structured AI prompts for NDA US analysis elevates routine contract review from a manual bottleneck into an efficient, repeatable workflow. By systematically querying the AI on definitions, standard carve-outs, statutory compliance under the DTSA, and choice of law, contract reviewers can quickly surface hidden operational risks. Build a library of verified prompts, apply them consistently across your agreement intake pipeline, and ensure a qualified legal professional performs the final review to protect your legal and business interests.