As artificial intelligence becomes more capable, legal teams are beginning to hear an uncomfortable question: If the business can put a contract into an AI tool, why does it still need human interference?
Brock Hubert, Deputy General Counsel at Highline Warren, has heard versions of that question from his own colleagues. Sometimes it is delivered as a joke, but usually there lies some seriousness behind it. For an episode of Squared & Settled, Brock sat down with Ashlyn Donahue, Senior Director of Legal at LinkSquares, to discuss what AI changes about the role of in-house legal — and what it does not.
AI can identify issues, summarize documents, propose language, and make information easier to access. But legal’s value has never come from producing redlines alone. It comes from understanding the organization, its priorities, its tolerance for risk, and what the person requesting help is actually trying to accomplish. That context is difficult to reproduce through a prompt.
Rather than making legal less relevant, AI gives legal teams an opportunity to become more influential business partners. To take that opportunity, lawyers need to understand the technology, help the organization use it responsibly, and show how legal judgment supports the outcomes the business is trying to achieve.
AI Cannot Replace Business Context
A generic AI tool can review a contract and identify provisions that appear unusual or risky. It can also return a redline with, as Brock put it, “nine trillion things on it” that may not matter to the company or the deal. The difference is context.
An in-house lawyer knows which risks the organization accepts, which provisions consistently require escalation, and where there may be room to compromise. They also understand that the purpose of contract review is not to produce the most comprehensive list of objections. It is to help the business complete an objective within appropriate guardrails.
That requires asking questions an AI system may not know to ask:
- What are the parties actually trying to accomplish?
- Which terms matter most in this relationship?
- Is the proposed language inconsistent with the commercial agreement, or merely different from the company’s standard language?
- Will pushing on this point protect the business, or delay the deal without meaningfully reducing risk?
As Ashlyn noted, a contract may return to the sales team covered in redlines while the parties remain conceptually close. Legal’s job includes explaining that distinction and helping the business understand what the edits mean.
AI may accelerate the review. It does not replace the institutional judgment that determines which changes matter.
AI Literacy Gives Legal a Seat at the Table
AI can feel like a magic box when someone first begins using it. Learning what it can and cannot do makes the technology less mysterious and gives lawyers a stronger foundation for evaluating tools built around it. That literacy also matters beyond legal’s own workflow.
When a business team proposes a new AI use case, legal needs enough technical understanding to ask informed questions. What information enters the system? How is that information processed? Where does it live? Who else can access it? Does the tool use the company’s data to improve its models?
Legal does not need to become the engineering department, but it cannot credibly guide AI adoption while treating the underlying technology as unknowable. This creates a unique opportunity. In many organizations, legal is not simply being called in to review a completed initiative. It is joining technology and business leaders earlier to help determine how AI should be used.
Brock does not believe legal needs to own that work alone. It does, however, need to be at the table.
Governance Requires Education, Not Prohibition
When ChatGPT first became widely available, some companies responded by banning AI use entirely. That position rarely lasted.
Employees can access public tools from personal devices. AI is also appearing inside software the company already uses, sometimes through updates that do not arrive through a formal procurement or legal review process. A blanket prohibition does little to address either reality.
“You’re not going to stop it,” Brock said. “The technology is too emergent.”
The more useful position is creating a framework in which people can use AI safely. At Highline Warren, a cross-functional AI committee began by assessing the company’s existing landscape. What tools were employees already using? What should they be using? What information could those tools access? What risks required company-wide guidance?
From there, the group developed an acceptable-use policy and began educating employees on the company’s approved tools, how data flows into them, and why using a protected company system is different from placing sensitive information into a public model.
That education cannot assume everyone begins in the same place. A workforce may include people who grew up questioning everything they encountered online and others who are less familiar with how generative AI produces information. Some employees may challenge an AI response instinctively, others may interpret a confident answer as a reliable one.
Effective governance accounts for those differences. It gives people enough understanding to recognize when an output requires verification, when a change to a tool should be flagged, and when the information being shared is too sensitive for the proposed use.
Data Is Part of the Exchange
The spread of AI through existing software creates another challenge: change does not always enter the company through formal channels.
A familiar SaaS provider may add an AI feature, update its privacy policy, or change how customer data is processed. An individual user may receive a pop-up notice and close it without realizing that the terms of the relationship have changed. This is not an entirely new problem. Software providers have always introduced new functionality, and a tool approved for one business use may later be adopted by another team handling more sensitive information.
AI makes the issue more visible because data is central to how many of these tools operate. Brock recommends helping employees think about data as a tangible, valuable asset. And the same principle applies to free tools. When a product costs nothing, the company may still be providing something of value in return. That value may be the information entered into the system, usage data, or permission to use content in ways the employee has not considered.
Legal cannot personally review every click or software update. However, it can teach people to develop the “spidey senses” to notice when something feels different and when to raise their hand.
Start With Work That Builds Confidence
Legal teams do not need to begin their AI adoption with the most complicated or ambitious use case. Brock recommends starting with work that is easy to understand, verify, and connect to an immediate benefit.
Research is one example. AI can help an in-house lawyer develop baseline knowledge before consulting outside counsel. Instead of paying an expert to explain an entire unfamiliar subject from the beginning, the lawyer can arrive with a working understanding, better questions, and a clearer sense of where specialized advice is required. That does not eliminate outside counsel, it helps legal spend outside counsel dollars more effectively.
Contract review offers another practical starting point. A legal team can place its playbook or drafting rules into an appropriate system (like LinkSquares) and use AI to flag deviations, screen for important clauses, or propose a first round of revisions. The lawyer still reviews the work, but the system can help catch the repeatable issues that should appear consistently across every agreement.
For Brock’s team, consistency is a central objective. The business should be able to anticipate the company’s positions rather than receiving a different answer depending on which lawyer reviews the contract. That predictability benefits more than legal. When sales or procurement teams understand the company’s standard positions, they can identify obvious conflicts earlier, prepare counterparties for likely edits, and send work to legal in better condition.
AI is useful here because it reinforces established judgment rather than trying to invent it.
Legal Can Lead Without Owning AI Alone
AI committees, acceptable-use policies, and contract tools may sound like operational projects. Together, they represent something larger for the legal department. They give legal a chance to change how the rest of the company experiences the function.
Brock described this as marketing the legal department. He does not mean creating an internal logo or sending promotional materials. He means using every interaction to demonstrate that legal is not simply a cost center or “the department of no.”
At Highline Warren, even changing the name of regulatory compliance to product compliance helped clarify the team’s purpose. The function exists not only to enforce rules, but to help keep products available, protect revenue, and move the business forward within established guardrails.
AI adoption offers another opportunity to demonstrate that role. Legal can help the company understand a new technology, identify risks without exaggerating them, and establish conditions that allow people to use it productively. It can turn something unfamiliar and potentially risky into a capability the organization can scale with greater confidence. That is more valuable than arriving at the end of the process with a list of reasons the company cannot proceed.
Better Partnership Starts With Understanding the Outcome
Brock has viewed the relationship between legal and the business from both sides. Before returning to an in-house legal role, he served as the CEO of a company and worked directly across sales, marketing, and operations. That experience reinforced a simple point: Legal and business leaders often make the same mistaken assumption about one another.
Legal may assume executives are ignoring risk and expecting the lawyers to clean up the consequences later. Executives may assume legal cares only about the legal issue and has little interest in the commercial outcome.
Neither assumption is accurate. Business leaders are often considering risk, even when they do not use the same language or reach the same conclusion as legal. Lawyers are often trying to help the company achieve its goal, even when they are raising concerns or asking for a different path.
Better business partnerships begin by assuming good intent. For legal, that means understanding what the partner is trying to accomplish before offering an answer. It also means explaining risk in practical terms. Citing a regulation is not the same as explaining what the regulation means for the decision, the deal, or the customer.
AI does not change that responsibility, it only makes it more important. As the technology takes on more of the repeatable work, legal’s value will become even more closely tied to context, education, and judgment. The opportunity is not simply to use AI faster than other departments. It is to help the organization make better decisions about where AI belongs, how it should be governed, and which outcomes it should support.
Legal teams do not become indispensable by protecting every task they perform today. They become indispensable by understanding the business well enough to guide what comes next.
To hear Brock Hubert and Ashlyn Donahue discuss AI governance, legal leadership, practical use cases, and the changing role of in-house counsel, watch the full episode of Squared & Settled.