AI and clause anomaly flagging at signature: last-minute review of late changes

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-legal-AI-and-clause-anomaly-flagging-adults

1. When an AI flags a change as 'ambiguous' during a late-stage document review, what should a reviewer conclude?

   1. The change requires human judgment to determine its impact
   2. The change is definitely harmless and can be ignored
   3. The AI has detected a formatting error
   4. The change violates a specific statute or regulation

2. Which type of document change would an AI most reliably flag as 'clearly substantive' during a signature-ready review?

   1. Reformatting section headings from numbered to bulleted
   2. A change in font from Arial to Times New Roman throughout the document
   3. An alteration to the indemnification cap amount in the liability section
   4. A correction of a typographical error in the notice provision

3. A junior associate relies solely on the AI's main-body diff and approves a contract for signature. The deal closes, and later a material adverse change clause in Exhibit C is discovered to have been weakened. What limitation of AI was illustrated here?

   1. AI cannot validate signature authority
   2. AI cannot replace a final read-through by counsel
   3. AI cannot detect changes to cross-references between documents
   4. AI cannot access cloud-based document management systems

4. During an AI-assisted signature review, the system flags a change in a defined term's capitalization from 'Material Adverse Change' to 'material adverse change' throughout the agreement. How should this be categorized?

   1. Clearly substantive, because defined terms must maintain consistent capitalization
   2. Benign formatting, because it does not alter the defined term's substance
   3. Clearly substantive, because it violates drafting conventions
   4. Possibly substantive, because it may indicate an unintended scope change

5. An AI flags changes in three places: the governing law section (word change), Schedule A (replaced an entire exhibit), and a footer element (page number format). Which location typically demands the most careful human scrutiny?

   1. Schedule A, because substantive terms often hide in exhibits and schedules
   2. The governing law section, because it affects where disputes are resolved
   3. The footer element, because page number errors can invalidate the entire document
   4. All three require equal scrutiny

6. A law firm uses AI to compare the signature page against the version counsel approved, but the deal falls apart when it emerges a side letter was executed that contradicts the main agreement. Why did the AI miss this?

   1. The AI was not sophisticated enough to detect verbal contradictions
   2. The side letter existed outside the four corners of the document
   3. The AI incorrectly categorized the side letter as a formatting change
   4. The AI was comparing the wrong versions

7. Which scenario best illustrates the appropriate use of AI in a signature-ready review workflow?

   1. Use AI to replace counsel review entirely for routine matters
   2. Use AI only for main-body text and skip the exhibits
   3. Use AI to flag changes, then have counsel read the full document with particular attention to flagged items
   4. Use AI to review the document and then sign without any human review

8. An AI system flags a change that modifies the notice period in the definition section from '30 days' to '30 calendar days.' Why might this be flagged as possibly substantive rather than clearly substantive?

   1. The change is obviously substantive because it adds specificity
   2. The change is clearly benign because both terms mean the same thing
   3. The change could be substantive if 'business days' was intended but was mistakenly left as 'calendar days'
   4. The change is clearly substantive because it lengthens the notice period

9. When reviewing AI-flagged changes, a senior attorney focuses only on those marked 'clearly substantive' and ignores 'possibly substantive' items. What is the risk of this approach?

   1. Only clearly substantive items can affect enforceability of the agreement
   2. The AI will automatically correct 'possibly substantive' items before signature
   3. Substantive changes could be hiding in the ambiguous category, leading to missed issues
   4. No significant risk; clearly substantive items are the only ones that matter

10. A legal operations team wants to implement AI signature review but is concerned about false negatives—situations where changes exist but the AI does not flag them. Which factor contributes most to false negative risk?

    1. Using the wrong comparison pair, such as comparing initial drafts rather than the final reviewed version
    2. The AI's difficulty with handwritten annotations
    3. The AI comparing documents that are too similar in content
    4. The AI's inability to read documents written in foreign languages

11. Why does the lesson specifically warn against letting the AI's main-body diff 'lull you into skipping the back of the document'?

    1. The back of the document is printed on different paper and cannot be compared electronically
    2. AI cannot read the back of the document due to technical limitations
    3. Substantive changes frequently hide in schedules, exhibits, and attachments
    4. The back of the document contains only administrative pages that require no review

12. During deal closing, the AI flags a change in the signature block from 'Authorized Signatory' to 'Vice President, Business Development.' Why might this warrant review beyond the change itself?

    1. The AI always flags signature block changes as substantive
    2. This change requires regulatory approval to be valid
    3. The change is irrelevant to the legal terms of the agreement
    4. The change could affect signature authority validation and may indicate an unauthorized signer

13. A client asks whether AI can be used to catch all late-stage changes and eliminate the need for a manual final review. What is the correct response?

    1. Yes, but only for documents under a certain page count
    2. No, AI cannot replace human judgment and must be used as a辅助 tool only
    3. No, AI is completely unreliable for legal document review
    4. Yes, modern AI can reliably catch all substantive changes with 99% accuracy

14. The AI diff shows that between the counsel-reviewed version and signature version, the word 'including' was changed to 'including but not limited to' in multiple locations. How should this be categorized?

    1. Irrelevant to the agreement's terms
    2. Clearly benign, because these phrases are interchangeable
    3. Clearly substantive, because this language broadens the scope of the provision
    4. Possibly substantive, because the change could be intentional expansion or a style edit

15. A firm implements AI review and discovers that most of the flagged changes in their recent transactions were categorized as 'benign formatting.' What does this suggest about their review process?

    1. The firm should stop using AI for signature review
    2. The AI is malfunctioning and requires technical support
    3. The firm should switch to a different AI vendor
    4. The document comparison may be catching minor changes that don't materially affect the agreements