Skip to content
Join LinkedIn group

Business & Practice Development

AT&T’s LegalEdge Shows How AI Could Change Outside Counsel Work

The competitive threat facing law firms may come from a client deciding it can perform more of the work itself.

The competitive threat facing law firms may come from a client deciding it can perform more of the work itself.

AT&T announced an OpenAI collaboration on October 8 built around LegalEdge, its internal legal-services operation. The company says LegalEdge handled nearly 100 matters during its first six months. Its stated approach keeps legal judgment, confidentiality and human oversight central to AI-enabled workflows. These are company-reported results and commitments. about.att.com

The commercial significance is where work goes - and which work remains with outside counsel.

Routine assignments are moving in-house

In an interview reported by Reuters, AT&T’s general counsel described moving many document-review, research, interview and first-draft assignments away from firms. LegalEdge has 38 professionals. He also suggested that approximately 30 national outside firms could become 18-20 over two years, without setting an exact target. Reuters

Those remarks describe one client’s strategy. They do not establish an industry-wide reduction in associate demand or prove that every internal assignment costs less.

They nevertheless provide a concrete example of how an AI-supported legal department can reconsider its purchasing decisions.

The client may arrive with more work already done

A traditional instruction often asks a firm to gather information, research the law, prepare a draft and advise on the result.

An expanded internal capability could change that sequence. Outside counsel might receive an existing factual record, internal analysis or proposed document and be asked to assess a narrower question.

That creates opportunities and complications.

A useful internal package could reduce duplication. An unreliable package could require substantial checking and reconstruction. The economic outcome depends on the quality of the handoff, the matter’s risk and the division of responsibility.

The presence of an AI-generated draft does not establish that the remaining work is simple.

Pricing must reflect the actual assignment

For Canadian firms, the useful question is what the client is buying.

If the client has completed part of the preparatory work, counsel should define the remaining scope carefully. Does the engagement include independent verification? Is the firm responsible for identifying missing facts? What assumptions can it reasonably accept?

A narrower mandate may support a different price. A verification-intensive mandate may require more work than the client expects.

Clear scope becomes essential to explaining the value of the engagement.

Junior development remains a management problem

Early research, review and drafting assignments can provide opportunities to learn. If clients commission fewer of them, firms must consider how junior lawyers will build the experience needed for more difficult work.

That does not require preserving inefficient tasks solely for training. It requires deliberate alternatives: supervised exercises, participation in strategy discussions, feedback on substantive decisions and appropriate responsibility on live matters.

Technology adoption and professional development need to be designed together.

AT&T’s announcement does not settle the future of the law firm. It sharpens a question firms already face: which capabilities make outside counsel worth retaining when the client can do more internally?

A stronger in-house team raises the standard for the work sent outside.

Related coverage: read our analysis of AI-supported legal service delivery and alternative fees.

Sources: about.att.com · Reuters.

Related coverage: Former Goodwin Partner Builds AI Law Firm Around Fixed Fees.