Ai for Lawyers and Legal Professionals

• A. A database that stores case law in chronological order
• B. Computer systems that perform tasks typically requiring human intelligence, such as language understanding, pattern recognition, and reasoning ✓
• C. Software that automatically files court documents without attorney review
• D. A physical robot that appears in courtrooms to assist judges

Why: AI refers to computer systems capable of tasks that normally require human intelligence, such as interpreting language, recognizing patterns, and drawing inferences. Simple digital storage or automated filing does not by itself constitute AI.

• A. Expert systems
• B. Large language models ✓
• C. Rule-based automation
• D. Predictive coding

Why: Large language models (LLMs) are trained on vast text datasets and generate fluent, human-like language. Expert systems rely on hand-coded rules, and predictive coding is a technology-assisted review method used in e-discovery.

• A. Because AI tools are legally prohibited from producing any citations
• B. Because AI tools can produce plausible-sounding but inaccurate information, including fabricated case citations ✓
• C. Because AI output is always copyrighted by the software vendor
• D. Because courts require every AI document to be notarized

Why: Generative AI can 'hallucinate,' producing confident but false statements, including nonexistent cases or misquoted holdings. Professional duties of competence and candor require lawyers to verify outputs against authoritative sources.

• A. The duty of competence, including understanding the benefits and risks of relevant technology ✓
• B. The duty to always charge hourly rather than flat fees
• C. The duty to avoid all use of computers in the office
• D. The duty to file every motion electronically

Why: ABA Model Rule 1.1, Comment 8, states that competence includes keeping abreast of changes in relevant technology. Lawyers using AI must understand its capabilities and limitations, and protect client confidentiality when inputting information.

• A. Technology-assisted review of large document collections during discovery ✓
• B. Delivering verdicts in jury trials
• C. Swearing in witnesses at depositions
• D. Serving process on opposing parties

Why: Technology-assisted review (TAR) and predictive coding are among the earliest and most widely adopted AI applications in legal practice, helping lawyers sort and prioritize large volumes of documents during e-discovery. The other options require human legal authority or physical action.

• A. Ask the AI tool to confirm that its citations are accurate
• B. Verify each case independently in an authoritative legal database or official reporter
• C. Check that the case names sound plausible and match the legal issue
• D. Rely on the tool's confidence score for each citation

• A. They retrieve cases from a corrupted legal database
• B. They generate text by predicting likely word patterns, not by looking up verified sources
• C. They deliberately invent cases to test the user's diligence
• D. Courts have not yet published enough decisions for AI to learn from

• A. Tell me everything about contract law
• B. Summarize the law on this topic and give me some cases
• C. In [jurisdiction], what are the elements of breach of fiduciary duty? Identify the governing authorities and flag any points where you are uncertain
• D. Is my client going to win this case?

• A. The summary will always be too long to be useful
• B. The summary may omit or misstate holdings, dicta, or qualifying language that changes how the case applies
• C. AI summaries cannot be saved or exported
• D. Summaries are only available for cases decided after 2020

• A. Replacing only the client's name while leaving other identifying facts intact
• B. Using a public AI tool without reviewing its data retention and training policies
• C. Anonymizing or generalizing sensitive facts and checking the tool's privacy terms before entering client information
• D. Assuming that any information typed into an AI tool is automatically protected by attorney-client privilege

• A. Replacing the lawyer's final sign-off on high-risk agreements
• B. Identifying and flagging specific clauses, such as indemnification, termination, or governing law, for human review
• C. Drafting binding legal opinions without attorney involvement
• D. Negotiating contract terms directly with opposing counsel

• A. Deleting outdated clauses from a contract template
• B. Automatically locating and pulling out defined provisions from a document so they can be compared or analyzed
• C. Translating a contract into another language
• D. Calculating the financial value of each clause

• A. AI tools cannot read PDF files
• B. AI may miss context, misinterpret ambiguous language, or produce confident but incorrect output, so a lawyer must verify results
• C. Courts require every contract to be read aloud by a lawyer
• D. AI tools are only legal for use on contracts under a certain dollar value

• A. Manually retyping each contract from a paper original
• B. Using templates with variable fields and logic rules to generate customized documents from structured inputs
• C. Scanning documents to create image files only
• D. Storing all contracts in a single unencrypted folder

• A. Uploading the contract to any free public chatbot for the fastest results
• B. Removing the client's name only, and leaving all other details in place
• C. Checking the tool's data handling and confidentiality terms and using approved, secure platforms before inputting client information
• D. Assuming all AI tools automatically delete data after each session

• A. Duty of confidentiality
• B. Duty of technological competence ✓
• C. Duty of candor to the tribunal
• D. Duty to avoid conflicts of interest

Why: Comment 8 to ABA Model Rule 1.1 states that competence includes keeping abreast of the benefits and risks of relevant technology. Lawyers using AI tools must understand how those tools work and what their limitations are.

• A. Whether the tool produces well-formatted output
• B. Whether the provider may retain, train on, or disclose the input data ✓
• C. Whether the tool is faster than traditional research
• D. Whether the client has paid the lawyer's fees in full

Why: Confidentiality duties under Model Rule 1.6 require lawyers to make reasonable efforts to prevent unauthorized disclosure. Lawyers must understand whether the provider stores inputs or uses them for model training, since that could expose client information.

• A. Charging the client for research time
• B. Failing to verify the accuracy of citations before presenting them to a court, implicating candor duties ✓
• C. Using a tool without the client's written consent for every task
• D. Failing to disclose the AI tool's pricing model

Why: Lawyers must not knowingly make false statements of law to a tribunal (Model Rule 3.3) and must provide competent representation. Presenting unverified, fabricated citations violates these duties, as several sanctioned attorneys have learned.

• A. Never, because AI is just a research tool like a database
• B. Only if the tool was developed outside the United States
• C. It depends; disclosure may be required when the use materially affects the representation or involves client information ✓
• D. Always, in every case, regardless of the task

Why: Model Rule 1.4 requires communication sufficient for the client to make informed decisions. Whether AI use must be disclosed depends on its significance and on whether client data is involved. Many jurisdictions and courts are issuing guidance on this, so lawyers should check current rules.

• A. Letting the AI tool work without any human review
• B. Ensuring that managerial and supervisory lawyers set policies and that staff use AI tools in a way consistent with the lawyer's ethical obligations ✓
• C. Recording every keystroke the AI tool makes
• D. Hiring a software engineer to maintain the AI tool

Why: Model Rules 5.1 and 5.3 require partners and supervising lawyers to make reasonable efforts to ensure that the firm and nonlawyer assistants, including technology, comply with professional obligations. This means establishing policies, training, and oversight of AI use.

• A. Collecting as much data as possible in case it becomes useful later
• B. Collecting only the personal data that is adequate, relevant, and necessary for a specified purpose ✓
• C. Storing all personal data indefinitely to support future audits
• D. Sharing personal data freely with third parties to reduce storage costs

Why: Data minimization requires organizations to limit personal data collection to what is necessary for the stated purpose. Collecting excess data increases privacy risk and regulatory exposure, so this principle is central to compliant AI systems that process client or personal information.

• A. Confirming the tool has a colorful and user-friendly interface
• B. Reviewing the vendor's data processing terms, including whether client data is used to train the model ✓
• C. Ensuring the tool can generate invoices automatically
• D. Installing the tool on every employee's personal device

Why: Lawyers owe confidentiality duties to clients. Reviewing vendor terms, including data retention, security, and whether inputs are used for model training, is essential to avoid unauthorized disclosure of privileged or personal information.

• A. A marketing review of how customers respond to a new product
• B. An evaluation of privacy risks posed by a processing activity and the measures to mitigate them ✓
• C. A financial audit of a company's annual data storage budget
• D. A background check on employees who handle data

Why: A DPIA systematically identifies and assesses privacy risks from high-risk processing, such as large-scale AI analysis of personal data, and documents safeguards. It is often required under the GDPR before deploying such systems.

• A. Anonymized data
• B. Aggregate statistics
• C. Personally identifiable information (PII) ✓
• D. Metadata summaries

Why: PII is information that can identify a specific person, such as names, addresses, or identification numbers, alone or when combined with other data. Recognizing PII is the first step in applying privacy safeguards to AI workflows.

• A. Anonymized data is always illegal to process
• B. Individuals may still be re-identified by combining the data with other available information ✓
• C. AI tools cannot process any form of anonymized text
• D. Anonymization only applies to financial records

Why: Data that appears anonymous can sometimes be re-identified through linkage with other datasets or contextual details. Lawyers must assess re-identification risk rather than assuming anonymization is a complete safeguard.

• A. Compare the tool's subscription price against the firm's annual technology budget
• B. Identify the specific legal tasks the tool will support and the risks those tasks carry ✓
• C. Ask the vendor for a list of its most famous clients
• D. Install the tool on all attorneys' devices to test it broadly

Why: Effective evaluation starts with the use case. Knowing which tasks the tool will handle, such as legal research, document review, or client communication, reveals the specific risks (confidentiality, accuracy, bias) that must be assessed. Price and vendor reputation matter, but they come after understanding the risk profile.

• A. A user interface glitch that displays incorrect formatting
• B. The model generating plausible-sounding but fabricated information, such as nonexistent case citations ✓
• C. The tool deliberately refusing to answer questions about sensitive topics
• D. A security breach in which outside parties access client data

Why: Hallucination refers to AI output that appears fluent and confident but is factually wrong or invented. In legal settings this can include fake case law or misquoted statutes, which is why lawyers must verify all AI-generated citations and authorities before relying on them in filings or advice.

• A. Whether the tool has a colorful and easy-to-use interface
• B. Whether the vendor's data handling, retention, and training policies protect client confidentiality ✓
• C. Whether the tool was recommended by a colleague at another firm
• D. Whether the tool produces output in under five seconds

Why: Lawyers have confidentiality duties under professional conduct rules. Before inputting client information, they must understand whether the vendor stores the data, for how long, who can access it, and whether it is used to train models. Ease of use and speed do not address these obligations.

• A. Accepting the output because the tool has a high accuracy rating in marketing materials
• B. Verifying key outputs against authoritative primary sources before relying on them ✓
• C. Asking the AI tool to confirm its own answers a second time
• D. Relying on the tool only for matters where the client is unlikely to complain

Why: Human verification against authoritative sources such as official reporters, statutory databases, and court records is the core safeguard. An AI tool checking its own output does not provide independent verification, and marketing claims are not a substitute for confirming the specific authorities cited.

• A. Vendors are legally prohibited from providing AI tools to law firms without one
• B. A policy sets clear expectations for how individual attorneys use AI, supervises staff, and documents oversight, reducing risk across the firm ✓
• C. Courts require every AI tool to be registered with the firm's internal policy office
• D. A policy eliminates the need for attorneys to review AI-generated work

Why: Vetting a vendor addresses one risk, but individual behavior creates others, such as employees pasting sensitive data into unapproved tools or relying on unchecked outputs. A written policy defines permitted uses, verification requirements, and supervision duties, and it helps the firm show reasonable efforts to meet its ethical obligations.

• A. Purchase the most expensive AI platform available
• B. Identify specific workflow bottlenecks and define the problems AI should solve ✓
• C. Replace all junior associates with AI tools
• D. Publicly announce the firm's AI adoption to attract new clients

Why: Effective AI integration starts with understanding where time, cost, or quality problems actually exist in current workflows. Choosing tools before defining the problem often leads to wasted investment and poor adoption.

• A. Assuming the AI vendor's terms of service automatically guarantee confidentiality
• B. Anonymizing client names only, while keeping all other case details in the prompt
• C. Reviewing the vendor's data handling policies and avoiding input of privileged information into tools that may retain or train on that data ✓
• D. Using AI only for internal memos so client data is never involved

Why: Lawyers have ethical duties of confidentiality under professional conduct rules. Firms should verify how vendors store, use, and train on input data, and avoid entering privileged information into tools without adequate safeguards.

• A. Hide AI use unless a client specifically asks about it
• B. Disclose AI use in a way that is transparent and consistent with engagement terms and any applicable ethical guidance ✓
• C. Disclose AI use only after a billing dispute arises
• D. Assume that disclosure is never required because AI is just a software tool

Why: Transparency supports trust and informed client decision-making. Many jurisdictions and bar opinions emphasize that clients should understand how AI affects their matters, including its role in work product and costs.

• A. To eliminate the need for attorney review of AI outputs
• B. To set policies on approved tools, permitted uses, supervision, and risk review across the firm ✓
• C. To reduce the firm's insurance requirements automatically
• D. To ensure that only partners are allowed to use AI tools

Why: A governance structure creates consistent standards for which tools are approved, what tasks they may support, how outputs are supervised, and how risks like hallucinated citations or data leaks are managed.

• A. It makes hourly billing completely obsolete overnight
• B. It may pressure traditional billable-hour arrangements, prompting firms to consider alternative fee structures and to bill for value rather than only time ✓
• C. It has no effect on billing because clients never ask about efficiency
• D. It requires firms to charge clients a separate fee for every keystroke

Why: When AI reduces the time needed for certain tasks, the billable-hour model can misalign with client expectations. Firms increasingly explore flat fees, value-based pricing, and transparent billing for AI-assisted work.