In May, the legal team at Okta flew to Montreal for its global offsite and ran a Shark Tank for AI ideas. It is the most concrete picture we’ve seen of AI for law departments in practice: mixed groups from every legal function pitching, a panel called “the council” judging, and the top three, chosen by team vote, presenting after one night of prep.
Kim Woodward, then VP of Legal and Chief of Staff at Okta, described the ritual on GC AI’s podcast, where senior legal leaders talk through how they’re running AI:
“It was like Shark Tank basically. We called the panel the council… the final three that were selected through votes, the top three, they came and did a presentation, they got one night to prep that together. Shockingly enough, most people, anyone who did a deck used AI.”
She spent the days before running around Montreal assembling gift bags of Canadian candy and maple syrup for the finalists, and the team committed to building the winning idea. Most of the pitches were designed to help the legal team broadly, because the groups had been deliberately mixed across practice areas.
That offsite is a law department embracing AI: competing to automate its own work, in front of its own leadership.
Running AI well across a law department takes an operating model, and the corporate legal teams pulling ahead built that first.
GC AI is the enterprise legal AI platform a three-time general counsel (Anki, Bloomtech, and Replit) built for in-house teams. As of August 2026, 2,000+ in-house legal teams use it, including 200+ public companies. Six legal AI features handle the department-scale work:
Playbooks: encode the department’s contract standards
Skill Library: shares working prompts across the whole department
Automations: handle recurring work like regulatory monitoring
Research: answers multi-jurisdiction questions with citations
GC AI for Word: drafts first versions where the redlines already live
Exact Quote: keeps every citation verifiable down to the character
How Law Departments Are Using AI in Practice
We’ve watched this shift since 2023: a handful of lawyers testing ChatGPT on their own, then dedicated legal AI platforms built for in-house work, and now shared department-wide infrastructure. That infrastructure means common playbooks, a shared prompt library, and a training cadence the whole team follows. The value compounds once the second, fifth, and fifteenth lawyer build on the same generative AI for legal fundamentals.
The pattern holds across 53 countries and customers like Columbia Sportswear, Jasper, Eventbrite, and Viant Technology.
The reach now extends past the lawyers, a preview of where department-wide AI is headed. Writing in Bloomberg Law in August 2026, GC AI CEO Cecilia Ziniti reported that roughly a third of GC AI platform users are non-lawyers in marketing, procurement, HR, and sales, and 80% of those non-lawyer users log in daily. A law department running AI well becomes “a platform on which the business runs.”
What In-House Law Departments Automate First
In more than 40 conversations with leading general counsels and legal operations leaders around the world, the first wave of automation lands in the same six places:
Contract review and redlining: NDAs, DPAs, and MSAs reviewed against the department’s own positions, the highest-volume work in nearly every law department. More in our guides to AI contract review and contract redlining software.
Legal research: multi-jurisdiction questions answered from primary law with citations, replacing the “quick question” that used to cost outside counsel a memo. More in our guide to AI for legal research.
Intake and triage: routing requests before a lawyer touches them. In a conversation on GC AI’s podcast, Marvell Technology’s legal ops team described auto-triaging about 40% of intake, with assignments landing within 24 business hours.
Compliance and regulatory tracking: standing automations that watch regulatory changes instead of a lawyer’s calendar reminder. More in our guide to AI for compliance monitoring.
First-draft generation: board consents, policy updates, and routine agreements drafted from the team’s own templates, inside GC AI for Word where the redlines already live.
Training the business: Trust & Will‘s senior counsel Alexandra Sepulveda uses GC AI to create NDA trainings for the sales team, themed “like a candy shop or a jungle adventure.”
How to Sequence the First Workflow
To start, think about which workflows repeat every week and how much time each one takes. Multiply the two, and for most departments that math points straight at contract review.
Set a baseline before the platform touches the work. A GC AI customer’s lawyer, at a publicly traded education technology company, put a before-and-after on the highest-volume workflow: “The prompt library is a game changer. NDAs used to take me 40 minutes. Now it’s 5 to 10.”
The spend numbers follow the hours. Another GC AI customer’s legal team, at a Fortune 500 IT solutions provider, reported: “Our legal bills dropped 30% in the first year.”
Run the first workflow for a month, log the delta, and bring that number to the second workflow on the list. The sequence compounds: each shared playbook or skill makes the next one faster to build.
Each of these starts as one lawyer’s workflow and becomes department infrastructure the moment it lives in a shared playbook or skill. The operating model exists to move each one from a private experiment into a shared standard.
The Operating Model: How Corporate Legal Teams Run AI
The strongest in-house law departments we work with run AI on three habits: a named owner, a training cadence, and visible metrics.
Give the Work a Real Owner
Woodward founded Okta’s legal ops function in 2018, after pitching a GC whose “first response was legal ops is not a thing.” By 2026 her team included a dedicated AI lead, and she puts the seniority bar directly:
“It cannot be, you know, a paralegal. I hear a lot of stories where it’s basically a paralegal who moonlights as your AI person on nights and weekends. That’s not the moment we’re in right now.”
The owner stays current on what the platforms can do, then translates that into the department’s workflows. In a five-lawyer department that owner is a fraction of someone’s role with real hours attached, and in a 15-person department it is a job title.
Put Training on the Calendar
Okta’s legal team made “two legal AI days a quarter” an annual KPI. Woodward describes how the first one landed:
“We did one that I would say landed with maybe the like most advanced people of the team, scared a little bit of the people in the middle and like lost everybody at the beginning. But it was kind of nice because it scared a lot of people.”
Her takeaway for the team: “everyone’s realizing that everyone’s behind. It really doesn’t matter where you are on the AI curve.” One line from the room stuck with her: “I don’t even know what an MCP is.”
Marvell’s legal ops team took the gamified route, running a year-long adoption push they called “the legally brilliant challenge.” The format matters less than the calendar entry. Practice builds the skill, and the calendar entry is how practice survives a busy quarter.
GC AI’s legal AI classes exist for exactly this gap. Taught by former general counsels, California CLE-eligible, and with 8,000+ lawyers taught to date, they give a law department a training cadence without building one from scratch.
If the first legal AI day needs an agenda, this one fits in an hour:
Pick one live matter: a real contract from this week’s queue, reviewed together against a playbook.
Show one saved prompt: a team member walks through a skill they use weekly and what it replaced.
Name the gap: each lawyer writes down one task they still do fully by hand.
Assign one build: the AI owner turns the most common gap into a shared skill before the next session.
Book the next one: the cadence is the point.
Make the Results Visible
Sharon Johnson, SVP and Chief Legal Officer at MODE Global, runs a six-lawyer department supporting a $3B transportation and logistics business at 100% AI adoption, and she tracks it:
“There are weeks I have saved over two headcounts for our team just using the tools, and I have KPIs to back that up.”
That is the number to bring to your next budget conversation. It sits naturally beside the corporate legal department metrics most teams already track.
That same lawyer called GC AI “my favorite product I’ve ever used as a legal professional in 30 years.”
AI-Native vs. Legacy Law Department Software
Law department software has meant tracking systems for two decades: a CLM for contracts, a matter management system for workload, e-billing for outside counsel spend, a document management system for storage. Those systems record the work. An AI-native law department platform performs the work: the review, the research, the drafting, the triage, with tracking running behind it.
Kerrie Forbes, Chief Legal Officer at JSX, runs both kinds of software:
“We’re setting up a scoreboard and a dashboard for tracking our matters and our spend and things like that. I don’t know that that helps so much with productivity. GC AI is probably what’s helped us the most with productivity.”
Here is how the same matter moves through each stack:
Stage | Legacy stack | AI-native department |
|---|---|---|
Intake | Email arrives, paralegal logs it in the matter system | Request auto-triaged, routed with context attached |
Review | Lawyer reads the contract line by line, checks the CLM for precedent | Playbook flags off-standard clauses against the team’s own positions in minutes |
Research | Outside counsel memo, or an hour in a research database | Cited answer from primary law, verified through Exact Quote |
Reporting | Quarterly export from three systems | Hours saved and matters closed, tracked as they happen |
The two categories complement each other: a system of record keeps the history, and the AI layer picks up more of the work itself. We map the full stack in corporate legal software and go deeper on the ops layer in our guide AI in legal operations.
What AI Can Save a Corporate Law Department
GC AI’s December 2025 ROI study of more than 100 active customer teams measured the department-level return directly:
14 hours saved per lawyer per week
14% reduction in outside counsel spend
Approximately $252,000 in annual savings for the median company
97.5% of teams see value before the end of month one
The $252K figure is the 14% reduction applied to the $1.8M median outside counsel spend reported in the ACC Law Department Management Benchmarking Report. Departments in the top spending quartile, at $11.2M+ in annual outside counsel spend, stand to recover even more.
Accuracy is the other half of the department math, because time saved on work that needs redoing saves nothing. GC AI’s In-House Legal Bench (May 2026), a benchmark GC AI built and ran across 100 in-house legal tasks with 1,200+ attorney-developed criteria, scored the platforms lawyers most often compare:
GC AI: 86.8%
ChatGPT (GPT-5.5): 79.8%
Claude (Opus 4.7): 68.4%
Gemini (3.1 Pro): 57.5%
GC AI’s widest margins came on regulatory tracking and legal research, two of the highest-volume categories of law department work.
Cameron Clark, Head of Legal at Arc’teryx, said it simply:
“With GC AI, we’ve handled the workload of a full legal team with just one or two lawyers.”
At Tipalti, Senior Director of Legal Joys Choi counts “609 hours saved this year alone.”
The team keeps the judgment calls: which risks to accept, when to escalate, when a deal needs a human no. Ziniti drew the parallel in Bloomberg Law: “Accounting software didn’t shrink the accounting field; it created strategic finance.”
What a Law Department Still Owns
These duties stay with the lawyers. A platform cannot own them for you. Four questions come up in nearly every rollout, and each one has a workable answer:
Verification: every AI output that leaves the department gets a lawyer’s review, and citation-level verification tools like Exact Quote make that review fast enough to hold. The State Bar of California’s practical guidance on generative AI treats output review as a core competence duty.
Confidentiality: run legal work through an enterprise platform under enterprise terms. GC AI is SOC 2 Type II and SOC 3 certified, GDPR compliant, and AES-256 encrypted, and every AI provider in its stack processes customer content on a zero data retention basis, documented on the live subprocessor list.
Privilege: how your team uses AI can surface in discovery, as the Heppner privilege ruling showed in 2026. Written AI-use practices protect the department’s position.
Ethics: ABA Formal Opinion 512 frames the duties of competence, confidentiality, and supervision for generative AI, and it reads as a checklist for a department rollout memo.
You stop being the bottleneck and start being the guardrail, as Bloomberg Law reported in August 2026.
How GC AI Fits an In-House Law Department
One general counsel we work with runs a ten-lawyer team at a private-equity-backed software company, and she skipped the CLM purchase entirely. Closed contracts run through a saved GC AI extraction prompt that pulls the terms into one table. A question like which customers hold a longer termination notice, which used to mean reading every contract, now takes a query.
Her company’s risk posture lives in the team’s Custom Company Profile, so a first-year contract manager’s redlines arrive calibrated to positions the GC already set. Her newer lawyers train with the platform too: they tell it they are new to the role, ask it to explain its reasoning, and bring what they learned to the group meeting each week.
The department-scale features are the same across these stories: Playbooks hold the contract standards, the Skill Library holds the working prompts, Automations watch the regulatory calendar, and Exact Quote keeps citations verifiable down to the character.
Shared standards, a named owner, visible results: the model works the same at ten lawyers as at two. The Shark Tank offsite is optional, and by the accounts of the teams that run one, worth the maple syrup.
One playbook, one owner, one number you can show the CFO. Your business moves fast. So should your legal team. Join the 2,000+ in-house legal teams using GC AI.
Frequently Asked Questions
What Is the Best AI Platform for a Law Department?
The best AI platform for a law department is one built for in-house work, with adoption and accuracy data behind it. GC AI meets that bar: purpose-built for in-house law departments, used by 2,000+ legal teams including 200+ public companies as of August 2026, and scored 86.8% on the In-House Legal Bench (May 2026), ahead of ChatGPT, Claude, and Gemini. The best legal AI tools for in-house counsel guide ranks the wider field by department fit.
How Much Does AI for a Law Department Cost?
GC AI starts at $500 per user per month on the Individual plan, with Team plans (including enterprise SSO and Solutions Attorney support) priced on request. Per-seat pricing is standard across legal AI, so a department can start with one or two licenses and expand on results. GC AI’s December 2025 ROI study found the median customer saves approximately $252,000 per year, 14% of the $1.8M median outside counsel spend in ACC benchmarking, with 97.5% of teams seeing value before the end of month one.
How Can Law Departments Use AI Without Compromising Confidentiality?
Run legal work through an enterprise AI platform under enterprise terms, with your own data governance policy on top. GC AI keeps each customer’s data in a segregated database, gives organizations control over which AI model providers are enabled, and logs who changed a model setting and when. Consumer AI accounts sit outside enterprise terms, so privileged work stays on the platform your organization governs.
What Should a Legal Team Assess Before Adopting an AI Tool?
Start with the pain point: name the workflow the platform must improve in its first month, and set the baseline hours it currently takes. Then review your organization’s data security and technology policies and bring IT and compliance stakeholders in early, since enterprise requirements decide the shortlist faster than feature lists do. GC AI’s 14-day free trial lets the team run the assessment on live work.
How Can Legal Teams Build an Internal Case for AI Adoption?
Anchor the case in numbers leadership already tracks: hours per lawyer per week, outside counsel spend, and matters closed. GC AI’s December 2025 ROI study of 100+ customer teams gives the benchmark figures (14 hours saved per lawyer per week, 14% outside counsel reduction), and a pilot on one live workflow turns those into your own. MODE Global’s legal team presents headcount-equivalent savings as formal KPIs.
