GC AI

Instacart Legal’s Nicole Altman & Kelly Noguchi Built an AI-First Culture That Stuck

Read time: ...

Most in-house legal teams are currently somewhere between ‘we watched a webinar about AI’ and ‘we have a Slack channel where people share AI prompts.’ Nicole Altman and Kelly Noguchi of Instacart Legal, are, by comparison, in another stratosphere. 

Nicole is Senior Counsel at Instacart, where she leads AI governance, privacy, security, and IP, plus builds AI tools that win hackathons in her spare time. Kelly is Senior Legal Technology and Operations Manager, where her job is to take everything Nicole and the rest of the team are building and figure out how to make it stick across a 60-person department.

Together, Nicole and Kelly have turned Instacart Legal into one of the most automation-forward legal departments in the country, and they’ve embedded AI workflows into every part of their day.

"It’s getting so good but it still doesn't replace us,” said Nicole. “AI can be a really amazing companion, assistant, and thought partner in how we do our work." 

In conversation with GC AI founder Cecilia Ziniti, Nicole and Kelly dished on the human side of scaling legal with AI, including the joy, the overwhelm, and the opportunities to lead change across the business.

AI-First Work Cultures are Built Top-Down

Years before most tech companies began introducing AI tools en masse, Instacart engineers developed an internal AI chatbot called ‘Ava’ on top of leading model APIs and made it available company-wide. Legal was the first to get access.

"You still hear of companies that don't have enterprise AI tools, where people are using shadow AI because the company isn't there yet,” said Nicole. “We're the opposite. We got enterprise AI very early on."

This AI-first mindset is an intentional part of Instacart’s culture and tech fluency is baked into personal and team goals across the company. But interestingly, the company’s leadership gave each team genuine leeway in how they got there. 

There were no mandatory workflows or one-size-fits-all rollouts. They made it clear that AI adoption was a priority, and gave people the time and space to figure it out. The result was a company full of professionals who use AI because it works, instead of out of obligation. Instacart’s positive culture made experimentation feel safe rather than risky, and empowered each employee to use their own judgement in how they used AI.

“I think AI is something that we all have to learn no matter what and incorporate into our day-to-day, but giving everyone the space to decide how they do that engages a lot more people,” said Kelly. “I feel like AI is just part of every conversation we have these days.”

Monthly business reviews have started including a standing segment where people from various parts of the legal team demo what they've been doing with AI. And when the team convened in person in San Francisco for their annual retreat, a major theme was cultivating an automation-first mindset. All of this was led and encouraged top-down by the company's General Counsel, Morgan Fong, which was critical.

Note from CZ: The contrast between Instacart and most legal teams I visit couldn't be starker. I've walked into departments where the GC is skeptical about automation and the team is waiting for permission. At Instacart, the permission was granted years ago and the team has been running fast ever since. Early leadership involvement is the key. 

How to Develop a Team of Early Adopters

Nicole is the living counterargument to the myth that AI adoption fails when it comes from the top. She’s spent her entire legal career getting hands-on, making mistakes, pushing through the frustration of using new tools, and doing it again the next week when the technology changed. 

She describes the typical AI adoption process like a friend who tried to write a condolence card with ChatGPT, got something too flowery, and gave up. They never thought to ask the AI to make it less flowery. The impulse to try again, to tweak the prompt, or approach from another angle is what makes an early adopter, and these types of thinkers tend to get the most value out of automation.

"It takes a lot of investment of time and grit and patience to get the good stuff out of these tools,” said Nicole. “You have to keep doing that over and over again because the systems are so rapidly changing. What they can and cannot do today is very different even from three months ago.”

Nicole and Kelly make the learning curve a little easier for Instacart’s employees by offering lots of opportunities to practice, including live demos, learning sessions, and a Slack channel where AI wins and discoveries get shared in real time. Their goal is to make curiosity the driver rather than fear.

"The best way to get people into using AI is to show concrete examples of what it can do and walk people through it, over and over,” said Nicole.

The Gap Between Personal and Enterprise AI Scaling

Kelly is one of the most honest voices in legal ops about how hard AI implementation is for large enterprises. She draws a sharp distinction between two very different AI use cases. 

The first is for personal productivity: using a chatbot to brainstorm, draft, research, or pressure-test an argument. That win is immediate, accessible, and deeply underrated. The second use case is building scalable enterprise workflows that the whole team can rely on. That level of automation requires change management, patience, and a tolerance for things not working on the first try.

"The gap between personal productivity and a scalable, business-ready process is where the huge challenge is,” said Kelly. “We can't glaze over actually changing excitement into execution. That's real." 

The infamous MIT study on AI implementation gets an honest read here. Kelly says that the study isn't proof that AI is overhyped, it's a snapshot of the exact transition point every legal team is in the middle of. The tools we have available are moving faster than any organization's ability to absorb them, and that gap is a people problem as much as a technology problem. It’s on your in-house leadership to teach everyone else how to incorporate automation into their workflow in a way that’s sustainable for the business. 

Note from CZ: In 1996, the Illinois State Bar sought public comment on whether email was ethical for lawyers to use. Chief Justice Roberts' 2024 report compared AI adoption to the word processor transition. We are living through a similar moment right now, and Kelly's honest accounting of its difficulty is exactly the kind of grounding we need.

How ‘Vibe Coding’ Helped Instacart Solve a Painful Problem

Nicole entered her first Bolt hackathon in part because Kelly gave a demo that finally pushed vibe coding to the top of her professional development list. She ended up building something that solves one of the most universally painful parts of contract management.

Every lawyer knows the frustration of searching for a counterparty in your contract database and getting back dozens of documents: the original agreement, amendment one, amendment seven, a restatement that only partially superseded the prior version. There is no single document that tells you what the operative terms are. Nicole’s new AI tool, Contract Compass, fixed that for Instacart Legal.

"It takes the various pieces of a contract, understands using AI what has been superseded, and gives you a change log and a final version,” said Nicole. “It’s GitHub for contracts." 

Kelly points to Nicole's as a great example of someone that's baked in AI into their life and it makes her a better teammate, a better attorney, and now, a competent vibe coder.

Note from CZ: In mid-2025, a GC AI user uploaded 26 documents spanning three spinoffs and a primary agreement dating to 1995. The AI gave back exactly what Nicole described: the operative version, the amendment history, in minutes. Today, you can do the same with GC AI Playbooks. Nicole built her version herself at a hackathon called Contracts Nirvana! Talk about taking matters into your own hands.

GC AI makes contract review simple, and you don’t even need to enter a hackathon! Easily review contracts against your company’s standards, turn your judgement into repeatable workflows, and save time while you’re at it with GC AI, your in-house legal team’s new secret weapon.

Listen to the Episode

Listen to the Episode

GC AI: Legal AI, for In-House

GC AI: Legal AI, for In-House

14 HRS

Saved per week per lawyer

21%

Greater accuracy than generalist AI

1,600+

In-house teams trust GC AI

GC AI scored 86.8% across 100 in-house legal tasks ahead of leading AI models

79.8%

ChatGPT (GPT5.5)

68.4%

Claude (Opus 4.7)

57.5%

Google Gemini (3.1 Pro)

GC AI led in every one of the 10 task categories, with the largest margins in research-intensive tasks

Back To Top

Back To Top

GC AI

Take the first step now.

Let’s explore about how we can make your life
as an in-house lawyer a whole lot easier.

Take the first step now.

Let’s explore about how we can make your life
as an in-house lawyer a whole lot easier.

Back To Top