As generative AI and large language models (LLMs) gain traction across the legal industry, the dynamics between in-house and outside counsel are shifting quickly. Many of our in-house clients are already leveraging AI to streamline routine work, reduce costs, and manage more matters internally. Still, one thing remains clear: AI cannot by itself replace the judgment, strategic thinking, and trusted guidance that experienced lawyers bring to the table. Rather than replacing the lawyer, AI is a tool that can enhance the attorney-client relationship.
In this shifting landscape, the firms that thrive will be the ones that pair AI-driven efficiency with the very human strengths clients still rely on most — practical counsel, sharp negotiation, and business-aligned risk management. Lawyers should remain flexible to incorporate AI tools as they continue to evolve in real-time.
Where AI Falls Short
Recently, one of our clients ran a reseller agreement through ChatGPT before sending it to us for review. The tool flagged the client's "unlimited liability" as a key area of concern, but that analysis was actually completely wrong. Upon reviewing the contract, we realized that – contrary to ChatGPT's confident-sounding conclusions, our client had not actually been exposed to unlimited liability; instead, their customer had capped its own liability without offering our client any reciprocal protection. This unbalanced allocation of risk was completely missed and mischaracterized. Worse, the tool overlooked more critical structural flaws in the contract, like whether key software vendor terms were properly flowed down, whether the reseller had adequate indemnity protection, and overall whether there were significant legal and business risks that could leave our client unnecessarily exposed.
These types of contractual issues pose real legal and business risks that only experienced lawyers can spot in context. So, while AI may be capable of initially flagging surface-level issues, it still can't perform higher-level functions, including:
- Understanding how different provisions work together
- Weighing competing risks
- Identifying what is missing
- Prioritizing risks based on the deal structure
- Crafting practical solutions that help close deals quickly
When AI Provides Faulty Output
We're also seeing clients run into pitfalls when AI-generated contract language looks plausible on its face, but falls apart under real-world scrutiny as the tools suffer from "hallucinations" in returning key output. In another recent example, when asked to generate a template for an agreement between a service provider and its customer, the AI produced a reasonable looking SaaS agreement, but upon review we found several "holes," —many standard provisions were missing. AI can also misinterpret the purpose of particular clauses, overlook key terms, and flag the same issue multiple times. Likewise, it may label clauses as "standard" when they are actually highly one-sided or inappropriate for your situation. We've even seen several documents including provisions that sounded like legal language to the non-lawyer client, but were actually meaningless when read more carefully and had to be scrapped and replaced with proper contract language.
In the above examples, we solved for these limitations by leveraging AI to save time (e.g. examining the client's input or prompt to home in on what the client needed), combined with our own experienced review to identify the risks and deliver practical business advice.
The Smart Play: AI + Trusted Legal Advice
In this way, we acknowledge that the "AI law firm of the future" is here today as long as the AI tools are used in a supportive role to supplement the lawyer's review – but not to replace it, and for lawyers to continue to evolve as these tools evolve. Our role as outside counsel is to help you capture all the upside of AI's speed and efficiency without taking on hidden risks that the tools can't see.
If your legal team is experimenting with AI (ChatGPT, Claude or any other GenAI tool) to work smarter and faster, we're here to support that momentum and not stand in its way.
Indeed, looping us in early, by sharing your prompt and the output, enables us to get oriented quickly, including:
- Understanding your priorities and concerns
- Identifying issues that may have slipped through AI's net
- Focusing your attention where it matters most within the context of the deal or industry at large.
- Delivering faster, more tailored guidance
We can also support your team's AI use by:
- Highlighting what AI can do (save time, brainstorm) and cannot do (identify risk, offer strategic counsel)
- Refining your prompting strategy
- Creating practical guidelines for how your team uses AI
- Building playbooks that integrate AI and legal counsel
- Providing legal judgment, commercial perspective, and negotiation strategy that only experienced counsel can deliver on top of the AI tools.
Accelerate with Confidence
At OGC, we blend business-minded legal judgement with deep experience across disparate industries and deal structures. We understand how legal terms play out in the real world — at the negotiating table, in commercial relationships, and when risk becomes real.
By thoughtfully integrating AI tools into our legal practice, we help our clients capture AI's advantages without sacrificing strategy and substance. We evaluate risk in context, identify what's missing, and ask the questions no algorithm can.
Our goal isn't to slow you down. It's to help you move faster and more confidently. That means knowing what to flag, what to ignore, and what deserves a closer look before you sign.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
