RBMA Bulletin Feature: Sign With Confidence: How AI Gets You There

September 3rd, 2026

By: Andrew Mims, Director of Finance

This article was originally published in the July/August 2026 edition of the RBMA Bulletin. View the article in its original format here. 

IN FOCUS

• AI Helps You Understand Before You Sign
• AI Strengthens Your Judgment, Not Replaces It
• Confidence Comes from Preparation

Understand What You’re Signing

At some point, almost every professional has a contract land on their desk that makes them nervous. It may be a new agreement from a potential partner, or a document that comes back from the other party’s attorney with more redlines than original content. Either way, it can feel like you need a dictionary and a second cup of coffee just to figure out what the legal jargon is really saying.

Contracts are part of doing business, but they are often written in dense language, and a single clause can carry a lot of meaning. A small word can change an obligation, or a vague phrase can create risk. We know the agreement matters, and it can be very frustrating when we do not fully understand what we are reviewing.

This is where artificial intelligence, or AI, can help. AI should never replace legal counsel, and it should never make decisions for you. It works best as an assistant to help you slow down, organize your thoughts, spot issues, and better understand the language before the contract goes to an attorney or decision-maker for review.

If used right, AI can make contract review feel less overwhelming. It can simplify complex provisions, highlight vague language, and help you assess whether the agreement protects the business in a fair and practical way.

The goal is not to turn all of us into attorneys, but to help us become more prepared and more confident when we review, draft, or negotiate contract language.

“AI works best when there are boundaries. The goal of imposing guardrails is to ensure AI supports good judgment rather than replacing it.”

Using AI to Help You Redline

I have found that redlining is often one of the most time-consuming parts of the contract process because it calls for precision. A good redline looks at what the other party’s language actually does. A single word or phrase can shift responsibility, narrow an obligation, or create ambiguity.

A great way to start is to provide the tool with a list of your priorities. This might include indemnification, contract terms, termination rights, service-level expectations, key performance indicators, insurance requirements, and more. The greater context you give it, the more valuable its suggestions will be.

Once you have established your priorities, you can begin feeding the contract sections into the AI tool and asking it to highlight language that conflicts with your priorities or exposes your company to risk. A great example is asking it to identify any clauses that shift liability in ways that disadvantage you, or to highlight language that gives the other party broad discretion without accountability. From there, you can also instruct the tool to suggest different language that better fits your intent while remaining reasonably fair.

Finding the Middle Ground

It’s important to remember that redlines should aim to protect your company without making the agreement unfair or unrealistic for the other party. It can be tempting to push every risk and obligation to the other side, but that will create unnecessary tension in your negotiation. A fair balance should place responsibility with the party best positioned to manage the risk.

AI can help with that fairness check. You can ask it to review proposed redlines and point out anything that may be too one-sided, difficult to negotiate, or disconnected from the actual risk. During contract negotiations, I find it helpful to ask for alternatives that still protect the organization but are more balanced and easier for both parties to accept.

Once AI understands your priorities, it’s great at suggesting revised language and explaining where the current wording may be too vague. Also, it can help explain why a proposed change matters from a business perspective, which is useful when discussing the issue with other invested parties.

You may come across provisions that feel incomplete but aren’t sure what was missing. AI can help you identify details on liability caps, deadlines, reporting requirements, notice obligations, service levels, remedies, exceptions, and termination rights. It can also help soften your initial redline that might sound too aggressive or tighten language that leaves too much room for interpretation.

What Did They Change?

When the other party returns a redlined agreement, instead of accepting or rejecting edits without entirely understanding them, a great process is to ask AI to compare the source language to the proposed revision and explain what changed and its impact.

Identifying shifts in liability, or sneaky language that quietly narrows your rights, becomes much easier when using a tool that can simplify the legal terminology. What might look like a minor rewording on the surface could eliminate a vital protection clause or introduce an ambiguity that works against you in a dispute. Use the tool to cut through the noise so you can walk into your next round of negotiations knowing exactly where you stand, what you are willing to accept, and where you need to push back.

AI as Your Drafting Partner

Most AI tools understand the general structure of common agreements, so they can help you create an organized first draft with the sections you would usually expect to see.

AI is not going to give you a copy and paste ready contract. The tool provides a structured framework to build upon. From there, you can apply your own judgment, find areas of concern that need to be stronger, and decide where the wording should create fairness.

Vague instructions lead to vague language. A draft is only as useful as the guidelines behind it.

The more specific you are about your priorities, the closer the first draft will be to something useful. Best practice is for you to know what the agreement is supposed to accomplish and what would create concern for your business before you begin.

Then, once AI creates your draft, don’t try to review the whole agreement all at once. You will forget most of your ideas by the end. Work through the draft section by section. Ask the tool to explain each one, identify what it’s trying to solve, and the intention behind it. Then ask whether the verbiage matches your original intent. AI can help you test whether the section is clear enough or whether it leaves room for interpretation.

Best practice when using AI for drafting is to treat it as both a drafting partner and an opposing receiver. Again, AI should not make the final legal decision, but it can help you create a stronger draft before it goes to legal counsel or another reviewer.

Important Guardrails

AI works best when there are boundaries. The goal of imposing guardrails is to ensure AI supports good judgment rather than replacing it.

A good place to start is by being honest with yourself about what AI can and cannot do. AI can summarize, compare, organize, and suggest language. What it doesn’t know is your organization’s risk tolerance, industry standards, operational reality, regulatory requirements, or negotiating history. Even if you provide that context, the output should still be treated as a draft, not the final solution.

Never forget that contracts carry real consequences that can impact your business for years. They can affect legal rights, financial obligations, exclusivity, and even future partnerships. AI may help identify an issue or suggest better wording, but real people still have to make the important decision of which risks are acceptable and whether the contract language fits the business relationship.

Confidentiality is one of the biggest guardrails to keep in mind. Never upload confidential agreements, protected health information, or other things like financial data, pricing terms, or other sensitive material without scrubbing the information in advance.

Additionally, every AI-generated suggestion should be checked against the actual contract and any applicable legal requirements. AI can appear confident even when it is completely wrong, so the output should always be treated as a starting point.

You also must test the language against operational reality. A clause may sound great and protective, but that does not mean the organization can actually comply with it. The entire contract needs to be practical for the people who must abide by it after it’s signed.

High-risk, complex, unusual, regulated, or heavily negotiated agreements should still go to legal counsel. The best use of AI in contract work is to prepare for that review by identifying questions, highlighting unclear language, comparing redlines, and organizing issues that require more examination. In the end, use AI to get organized, not to make the final call.