> ## Content Index
> Fetch the complete content index at: https://wire.fourthweb.ai/llms.txt
> Use this file to discover other available public pages before exploring further.

# AT&T Built an AI That Does What 200 Lawyers Used To
- URL: https://wire.fourthweb.ai/at-t-built-an-ai-that-does-what-200-lawyers-used-to/
- Published: 2026-10-24T18:30:47.000Z
- Updated: 2026-10-24T18:30:49.000Z
- Description: The lawyers aren't getting fired — they're getting cloned. AT&T's in-house legal team built LegalEdge with OpenAI to automate legal intake and scale attorney judgment across the company
- Author: Travis Wright
- Tags: AI Agent Economy, OpenAI

**The lawyers aren't getting fired — they're getting cloned.**

### The Summary

- [AT&T's in-house legal team built LegalEdge with OpenAI](https://openai.com/index/att-legal?ref=wire.fourthweb.ai) to automate legal intake and scale attorney judgment across the company
- Instead of replacing lawyers, the system acts as a force multiplier, triaging routine questions and freeing legal teams for high-value work
- Early signal that legal automation is moving from document review to actual legal reasoning at enterprise scale

### The Signal

AT&T has 160,000 employees generating legal questions every day. Before LegalEdge, most of those questions hit a bottleneck: the intake process. Business teams would email legal. Legal would triage. Simple questions that took 30 seconds to answer still took days to route. Complex issues got buried in the queue. [The company partnered with OpenAI to build a system that doesn't just answer questions](https://openai.com/index/att-legal?ref=wire.fourthweb.ai), it thinks like an attorney.

LegalEdge does two things the legal world has been trying to automate for decades. First, it handles intake at machine speed. Employee asks a contract question, the system pulls relevant precedent, checks against AT&T's internal policies, and delivers an answer with citations. Second, and this is the part that matters, it scales legal judgment. Not just retrieval. Not just search. Actual analysis of whether a proposed action creates risk.

> "The system acts as a force multiplier, triaging routine questions and freeing legal teams for high-value work."

This is what legal automation looks like when it actually works:

- Routine questions get answered in seconds instead of days
- Attorneys spend time on complex negotiations and strategy, not email triage
- Legal knowledge that used to live in partner-level heads becomes accessible across the org

The model runs on [OpenAI](https://wire.fourthweb.ai/tag/openai/)'s infrastructure but is trained on AT&T's internal legal corpus. That means it knows company precedent, internal policies, regulatory constraints specific to telecom. It's not generic legal advice. It's AT&T's legal department, compressed into an API.

Here's why this matters beyond one company's legal team. AT&T is a 143-year-old institution with deep risk aversion and a legal culture built on careful review. If they're comfortable letting an AI system handle initial legal analysis, the floodgates are open. Every enterprise with a legal team over 50 people is now asking the same question: what percentage of our legal work can we automate?

The answer is probably 60-70% of intake and initial analysis. That doesn't mean 60-70% headcount reduction. It means attorneys stop being glorified routers and start being strategists. The junior associate who used to spend 40 hours a week on contract review now spends that time learning to negotiate. The mid-level attorney who answered the same compliance question 50 times a year now builds frameworks for new product launches.

### The Implication

Watch for two things. First, a wave of similar announcements from other Fortune 500s in the next six months. Legal departments have been waiting for someone else to go first. AT&T just opened the door. Second, a spike in demand for lawyers who can train and audit AI systems. The skill isn't legal research anymore. It's legal architecture — building the judgment layer that agents execute.

If you're in-house counsel, this is your opening. The firms that figure out how to 10x attorney productivity with AI will own the next decade. The ones that insist on billing by the hour will become boutiques.

### Sources

[OpenAI Blog](https://openai.com/index/att-legal?ref=wire.fourthweb.ai)