AT&T and OpenAI: What This Means for In-House Legal Work
Josh Bertini, Head of Growth•Published
On October 8, AT&T announced that its legal department will work with OpenAI as a “flagship in-house legal design partner” to design, build, and evaluate AI workflows for AT&T's lawyers.
In January 2026, AT&T built LegalEdge as its own in-house team of attorneys, paralegals, technologists, and legal operations and e-discovery specialists, focused on AI-assisted legal work. AT&T calls it an “AI-first center of expertise,” and it handled nearly 100 matters in its first six months.
In-house lawyers know the company's business, risk tolerance, and history. AT&T's partnership with OpenAI aims to make that knowledge easier to use across the company, while lawyers decide how AI supports the work. For other legal departments, the opportunity is to turn their own expertise into workflows colleagues can use and lawyers can review.
GC AI is a turnkey legal intelligence platform built for in-house counsel, used by more than 2,100 legal teams, from solo GCs to Fortune 500 legal departments. With GC AI, your team turns its negotiation positions into contract redlines, finds obligations across signed agreements, and answers the company's legal questions with sources to check:
-
Spend your time on the calls that need a lawyer: Apply your approved positions to contract reviews and check cited legal research, so you can focus on exceptions and trade-offs.
-
Skip the platform upkeep: We build and maintain the legal workflows, and attorney-led classes get your team up to speed.
-
Budget with confidence and keep data protected: Refine prompts, revise documents, and keep asking follow-up questions without increasing your chat bill. Your company's information stays ring-fenced from other customers and is never used to train our model providers' AI.
What AT&T and OpenAI Announced
The partnership has three goals:
-
Making institutional knowledge easier to access
-
Handling appropriate routine legal work more efficiently
-
Giving legal professionals more capacity for “ideation, strategy, and advocacy.”
Lawyers decide how AI agents are used. An agent may answer a frequently asked question directly, or gather and synthesize information so a lawyer can give the advice. AT&T says governance, confidentiality, accountability, and human review will be built into every workflow.
AT&T plans to evaluate the work through quality, access to knowledge, reusable capabilities, and capacity for legal judgment.
What the Partnership Means for In-House Legal Work
AT&T announced that its corporate legal department would become a flagship in-house legal design partner for OpenAI three weeks after OpenAI launched Astra for Law on September 17. Those workflows are still in development.
Make Company Knowledge Easier to Reuse
David McAtee, AT&T's senior executive vice president and general counsel, put it this way:
“AI gives legal departments an opportunity to do more than improve individual productivity. It gives us an opportunity to rethink how legal work moves through an organization, how institutional knowledge is retained and how our lawyers focus their time.”
An answer in one lawyer's chat may never reach the next person handling the same issue. Shared positions, source documents, and reviewed answers give the team a starting point it can reuse. For example, a lawyer reviewing a supplier's renewal terms can start with the team's approved position, then assess any exception in the agreement.
Give Lawyers More Time for Business Decisions
McAtee wants legal's “knowledge and expertise more accessible to every AT&T organization we serve.”
On CZ and Friends, GC AI CEO and three-time general counsel Cecilia Ziniti recalled advice from Melissa Chaw, then a vice president and deputy general counsel at Amazon:
“You are a business person with a legal skill set.”
Start with the questions your colleagues ask repeatedly. Decide which they can answer using approved guidance and which need a lawyer to weigh the business trade-offs.
Decide Which Work to Keep In-House
For legal departments, keeping more routine work in-house is one possible result of putting company knowledge to work. In our December 2025 survey of more than 100 active customers, respondents reported a 14% reduction in outside counsel spend.
Our guide to reducing outside counsel spend with AI covers how to bring work back without overloading your team.
Account for the Full Cost of Building Legal AI
Keeping more work in-house also means deciding who maintains the systems behind it. A custom setup gives your team responsibility for configuration, integrations, testing, and updates. A maintained legal AI platform puts that platform work with the vendor; your lawyers still own company positions and legal review. Compare those responsibilities alongside subscription fees, research access, and any usage charges for the same workload.
On usage-based plans, long documents, repeated revisions, heavier reasoning, and extra model calls can increase consumption. Spending caps can control costs, but reaching a cap may interrupt work. GC AI Team plans include unlimited chats, so lawyers can refine prompts and ask follow-up questions without per-chat charges. Include separately priced services and your team's review time in the full comparison.
Start with a recurring matter and identify which parts your lawyers should own and which a platform can provide.
If your company already runs on ChatGPT or Claude, weigh how much of that work your legal team wants to take on itself.
Put Your Company's Legal Knowledge to Work With GC AI
GC AI brings your company's positions, source documents, and legal research into built-in workflows for your team. Your lawyers can review the supporting material, assess exceptions, and decide what advice to give the business.
Example: Review an NDA With Your Approved Positions
A sales lead sends an NDA for legal review. Apply your approved positions through Playbooks in GC AI for Word, then review the proposed tracked changes.
Use Exact Quote to inspect the contract language supporting a finding. The lawyer decides which exceptions the company can accept and approves the version to send. Record deal-specific exceptions with the matter. Update the shared playbook when the team approves a change to its standard position.
At Snyk, an AI security platform serving over 4,500 customers, including Google and Anheuser-Busch InBev, Senior Managing Counsel Alexis Palmer supports commercial and partner contracts and marketing reviews. She describes the balance:
“It doesn't replace the human element of risk assessment, it just gives me the time to focus on that part.”
After signature, GC AI's Contract Intelligence answers questions across your signed agreements with citations, such as which supplier agreements renew automatically, so the lawyer can open the cited clauses and confirm the notice deadlines.
Use GC AI in Your Existing Workflow
GC AI develops and maintains its legal workflows; your team keeps its company context and negotiation positions current. GC AI selects models across Anthropic, OpenAI, Google, and xAI, with fallback when a provider has an issue. Your lawyers review the resulting work before relying on it.
Our ChatGPT plugin connects your GC AI workspace to ChatGPT, so your lawyers can use the team's documents, playbooks, and standards from a ChatGPT conversation. The plugin is a separate integration from Astra for Law.
Your company's information stays ring-fenced from other customers and is never used to train our model providers' AI. We are SOC 2 Type II and SOC 3 certified and GDPR compliant.
More than 8,000 lawyers have completed GC AI Classes. Live classes cover practical legal workflows, with select classes eligible for California MCLE credit.
In our survey of 200 in-house legal professionals, 97.5% reported meaningful value before the end of their first month.
Jenna Hunt, then head of legal operations at Tipalti, now a GC AI Solutions Advisor, described the team's rollout:
“It didn't even feel like an implementation. It was just an on switch.”
See How It Works on Your Own Agreements
Bring one agreement your team reviews regularly, your approved positions, and a question the business needs answered. In a GC AI walkthrough, see how your team can review proposed changes, check the supporting sources, and decide what to send back.
