AI and the Next Generation of Business Lawyers: Using It for Good

By Trenton Roberts and Julia Ngo

Lawyers who are starting their careers now will be the first generation of business lawyers to spend every year of practice alongside generative AI. In a transactional practice, that matters more than most people outside the profession realize. Much of the work is high volume and pattern based: reviewing nondisclosure agreements, building diligence summaries, marking up commercial contracts, checking a purchase agreement against a term sheet, and keeping a company's governance records in order. That is exactly the kind of work these tools are getting better at.

The question for the next generation of business lawyers is not whether to use AI. It is how to use it for good: for clients, for the quality of the work, and for their own development as lawyers. Another of our firm's articles looks at how AI may change contract negotiation and why the negotiator's judgment becomes more central. This piece looks at the people doing the work. What should a junior business lawyer do with these tools, what should a firm teach them, and what rules keep the whole thing honest?

What the next generation of business lawyers can do with AI

Start with what the work looks like in a business law practice that serves companies, private equity sponsors, and private credit lenders.

A private equity sponsor evaluating targets may need dozens of NDAs turned around in a week, each one checked against the sponsor's positions on residuals, non-solicits, standstills, and term. We describe how that kind of program works in our post on high-volume NDA review. An acquisition generates a data room full of contracts that need to be read for assignment restrictions, change of control triggers, and exclusivity. A credit facility comes with a stack of ancillary documents that have to match the credit agreement. A growing company with a fractional general counsel sends a steady flow of customer agreements, vendor terms, and policy questions.

In each of those settings, a well-used AI tool can help with the first pass:

1. Issue spotting. Flagging the clauses in a counterparty draft that depart from a client's standard positions, so the lawyer starts with a map instead of a blank page.

2. Consistency checks. Comparing defined terms, cross references, and dollar thresholds across a set of related documents.

3. Summaries. Producing a first draft of a diligence summary or an issues list for the lawyer to verify and rewrite.

4. Drafting starting points. Suggesting alternative language for a provision so the lawyer can see options before choosing one.

The rule that governs every item on that list is simple. The lawyer verifies everything. The Professional Ethics Committee for the State Bar of Texas put it directly in Opinion 705, issued in February 2025: lawyers are responsible for the work product they submit regardless of who, or what, did the original research and drafting, and AI output can be a starting point but must always be carefully analyzed for accuracy and quality. The committee added that the lawyer's duty is more than catching false results. The lawyer is responsible for making sure the content is accurate and supports the client's interests.

That last point is where a young lawyer adds value. A tool can flag that a residuals clause is broader than the client's standard. Only a lawyer who understands the client's deal strategy can decide whether that clause matters in this process, with this counterparty, at this stage.

Efficiency that shows up for the client

The second way to use AI for good is to make sure the client benefits when the work gets faster.

The ethics opinions leave no doubt about this. The American Bar Association's Formal Opinion 512, issued July 29, 2024, says that a lawyer billing hourly may charge only for time actually spent. Texas Opinion 705 reaches the same conclusion: a lawyer may bill for the actual time spent using a generative AI program, including refining and checking its output, but may not charge hourly fees for time that was saved. Formal Opinion 512 also says a lawyer generally should not bill a client for time spent learning a tool the lawyer will use regularly.

Fixed fees raise a related point. Business clients often want predictable pricing for repeatable work, and fixed fees for defined tasks can serve them well. Formal Opinion 512 notes that if a tool lets a lawyer finish a task much more quickly, charging the same flat fee as before may be unreasonable. In Texas, Rule 1.04 of the Texas Disciplinary Rules of Professional Conduct prohibits unconscionable fees and lists the factors used to judge whether a fee is reasonable, whatever its structure. The healthy response is to let efficiency show up in the price and the turnaround, and to explain the basis of the fee clearly at the start.

Lawyers early in their careers are well placed to think this way. They have not spent decades measuring their value in hours. If they learn from the start that their value lies in accuracy, speed, judgment, and predictable cost, they will build practices that clients trust.

More time for judgment, negotiation, and client context

The most promising change for junior lawyers is in how they spend the day.

When the mechanical part of a task gets shorter, the time saved can go to the work that actually builds a business lawyer:

1. Understanding the client's business. Why does this sponsor care about the non-solicit more than the standstill? What does this company's sales team actually need from its customer agreement? Those answers come from conversations, not from documents.

2. Negotiation. Joining calls with counterparties, hearing how positions are traded, and learning when to hold a point and when to concede it.

3. Judgment calls. Deciding which of twenty flagged issues are worth raising, and how to explain the three that matter to a client who has five minutes.

4. Deal context. Learning how the pieces of a transaction fit together, from the letter of intent through closing, as we describe in our post on the middle market private equity deal process.

None of that happens automatically. A firm has to decide that time saved by a tool goes to development, not just to more volume.

How AI changes training and mentorship

Business lawyers have traditionally learned by doing the first draft and having a senior lawyer mark it up. The markup teaches. If AI produces the first draft, a new lawyer can lose that lesson and never develop the judgment to tell whether a draft is right. A lawyer cannot take responsibility for work they cannot evaluate.

That makes training more important, not less. A few approaches that make sense for a transactional practice:

1. Learn the positions before the tool. A junior lawyer should understand why a client's NDA playbook says what it says, as covered in our post on NDA negotiation points, before relying on any tool to apply it.

2. Draft first, then compare. Writing a provision from scratch and then reviewing alternatives builds skill. Starting from a generated draft every time does not.

3. Explain the markup. When a supervising lawyer changes a junior lawyer's work, the reason matters more than the edit. That conversation is the part of training a tool cannot replace.

4. Label the source. Formal Opinion 512 suggests marking materials produced by AI tools as such when they are stored in a client or firm file, so later readers understand where the work came from and how far to trust it.

The opinions also put responsibility on the supervising lawyer. Formal Opinion 512 says managerial and supervisory lawyers must make reasonable efforts to ensure that the lawyers and staff they supervise comply with their professional obligations when using these tools, and that training should cover how the technology works, its limits, the ethical issues, and secure handling of client data. In Texas, Rules 5.01 and 5.03 address the responsibilities of supervising lawyers for other lawyers and for nonlawyer assistants. Opinion 705 compares relying on AI to relying on an inexperienced or overconfident assistant. Mentorship is how a firm meets those duties in practice.

The guardrails that keep it honest

Using AI for good means using it inside the rules. In business law, five guardrails come up most often.

Competence

Comment 8 to Rule 1.01 of the Texas Disciplinary Rules says lawyers should strive to remain proficient in the benefits and risks associated with relevant technology. For a junior lawyer, competence means knowing both the law and the limits of the tool well enough to catch its mistakes.

Confidentiality

Transactional work runs on confidential information: target financials, deal terms, borrower data, and the very existence of a potential deal. Rule 1.05 protects that information. Opinion 705 advises lawyers to understand how a tool stores and uses what is entered, review its terms of service and data security, and train lawyers and staff on safe use. Formal Opinion 512 says that before entering client information into a self learning tool, a lawyer needs the client's informed consent, and boilerplate language in an engagement letter is not enough. Some clients also have outside counsel guidelines or confidentiality obligations of their own that address AI use, and those have to be followed.

Supervision

Every output needs a lawyer who reviews it and stands behind it. The supervising lawyer is responsible for building a process where that review actually happens.

Billing

Bill for actual time, keep fixed fees reasonable, and do not charge clients for learning the tools. Formal Opinion 512 also addresses when a tool's cost can be passed through as an expense and when it should be treated as overhead.

Candor

Formal Opinion 512 says lawyers must disclose their AI practices when a client asks, when the client's engagement terms require it, and in some situations without being asked, such as when the tool's use bears on the basis of the fee or when its output will influence a significant decision in the representation. Candor extends to the other side of the table too. Rule 4.01 prohibits a lawyer from knowingly making a false statement of material fact or law to a third person, and that duty does not change when a tool drafted the first version of the statement. For the occasional matter that reaches a court, Opinion 705 notes that some courts, including the Northern District of Texas, have local rules on disclosure of generative AI use in briefs.

What the next generation of business lawyers can build

If the lawyers entering practice now get this right, business clients will see faster first turns on routine documents, more consistent application of their positions across hundreds of agreements, and pricing that reflects the work actually required. The lawyers themselves will reach real competence sooner, because they spent their early years on judgment and client context rather than on formatting and repetition.

That outcome depends on young lawyers who are curious about the tools, skeptical of their output, and committed to the duties that come with a license, and on firms that train and supervise them with the same care. That combination describes a strong business lawyer in any era.

How we can help

We work with growing companies, private equity sponsors, and private credit lenders on NDAs, transactions, contracts, entity matters, and fractional general counsel needs, and we think carefully about how the work is done and priced.

If you want to talk about your legal needs, email info@robertsandwillie.com or call (469) 910-4017. Sending a message does not create an attorney-client relationship.

This article is for general information and is not legal advice. Attorney Advertising. Roberts & Willie, PLLC, Dallas, Texas; Trenton Roberts, responsible attorney.