PLI Legal Ops Institute 2025 - Panel 1💡 Real-Life Use Cases for Generative AI in Legal: Lessons from the Front Lines

A diverse group of in-house and firm leaders shared how they’re turning AI from theory into daily practice.

The recurring theme is AI success in legal is about thoughtful design, responsible deployment, and sustained stewardship.

🔍 The Takeaway

AI in legal is no longer experimental—it’s operational. The leaders driving it forward aren’t just deploying tools; they’re reshaping how legal teams approach technology, work, and deliver value.

🧭 Strategy First: Build, Buy, or Blend

Ashley Miller (CapGemini) set the tone: “If you don’t have the resources to enhance and maintain what you build, then buy.” She underscored the importance of selecting a tool fit to purpose and documenting the business requirements before selecting a tool.

Penny Williams (Google) emphasized grounding general-purpose AI tools in legal context: “Take the general tools in your arsenal and ground them for legal.”

Megan McMillin (Cleary Gottlieb) described how her firm’s development platform is reshaping their approach to emphasize design and development. She noted, “Product management isn’t optional—it’s essential.”

Jessica Williams (Geico) offered a pragmatic lens: “Before buying new tech, ask—what do we already have that can solve this problem today?” Only if the answer is “no” should one go to market.

⚖️ General vs. Legal-Specific AI: Finding the Balance

The panelists converged on one point: legal work has unique risk, nuance, and confidentiality requirements. Key take-aways:

✅ Carefully curate both tools and data.  If you start with a general tool, you need to match to your use case.
✅Select a legal-specific tool if it provides a better match to your company's control and security requirements.
✅Focus on the culture shift and need to build in feedback loops to guide adoption.
✅The real work is ingesting data from multiple sources, cleaning the data and pointing it to the right place.

Curation, governance, and trust supersede prompt engineering as tools advance.

🧰 Practical AI in Action

The examples shared were practical and specific:

• Ashley Miller uses Copilot to polish emails and memos, and to organize presentation thoughts — “AI elevates tone, clarity, and flow”—and to summarize contracts and executive briefings.
• Jessica Williams leverages AI to translate legalese into plain language and compare two contracts side by side.  Her team has also trained an AI agent on internal spreadsheets to match the right expert attorney in their panel to new matters.
• Megan McMillin is about to pilot FlashDocs for PPT generation and uses Legora for nuanced transcript analysis—while personally using AI to create outlines for work, as well as Relay to route kids school forms and Mantis to to pull together topic digests into a custom website every week.
• Penny Williams uses Notebook LLM to assemble first rough drafts from designated authoritative resources and to generate summaries, TL;Drs or briefing podcasts: “AI helps me move faster from research to insight.”
• Meredith Gordon built an outline for a hackathon, deck and to sort randomly for teams.

⚡ Lightning Round: Practical Advice

Closing tips were simple:

Penny Williams: “Use AI as an excuse to get organized and take advantage of public training resources .”
Megan McMillin: “Just start. Make AI a habit. Take the mindset of try something then iterate.”
Ashley Miller: “Be curious.”
Jessica Williams: “Ask AI how to write a good prompt.”

To view the PLI Legal Operations Institute 2025 recording, visit: https://www.pli.edu/programs/legal-operations/416133 (requires paid registration if your team does not have a subscription).