OpenAI’s latest move into the legal sector signals a shift in how enterprise AI may be deployed: rather than simply giving lawyers access to powerful models, the company is building an ecosystem around implementation, training and workflow integration.
Its newly announced partnership with London-based startup Telon illustrates that strategy. Telon, founded in June 2026 by former trial attorney Lewis Bretts, specializes in what it calls “legal engineers.”
These professionals are primarily former lawyers who help law firms and corporate legal departments configure AI systems, develop prompts and agents, and train employees to use the technology effectively.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
OpenAI has designated Telon a “select partner,” bringing the startup into its broader Partner Network. The significance of the arrangement lies in the problem it attempts to solve. Artificial intelligence has already demonstrated its usefulness for legal research, document analysis, drafting and other knowledge-intensive tasks.
Yet purchasing an AI subscription does not automatically transform a law firm’s workflow. Lawyers still need to understand how to structure requests, evaluate outputs, integrate AI into existing systems and maintain appropriate human oversight.
That implementation gap is becoming an important battleground in enterprise AI. OpenAI’s own work with law firms illustrates the opportunity.
Australian firm Gilbert + Tobin has reported using ChatGPT and Codex across its operations, reducing some recruitment research and data-extraction work from roughly four hours to 20 minutes and cutting selected conflict, KYC and AML checks to about five minutes.
The firm says governance and human accountability remain central to its deployment. OpenAI has also collaborated with Willkie Farr & Gallagher on firmwide AI adoption and the development of proprietary AI platforms through the firm’s innovation organization.
These examples suggest that the legal market is moving beyond experimentation toward deeper integration of generative AI into professional workflows. Telon’s model adds another layer. Instead of building another standalone legal AI application.
It positions people with legal expertise between the technology and the organizations using it. That approach could prove important because legal work carries unusually high requirements for accuracy, confidentiality and professional judgment.
A lawyer cannot simply accept an AI-generated answer because it sounds convincing. Legal conclusions must be supported by authoritative sources, relevant facts and applicable law. Errors can create financial, regulatory and reputational consequences.
The human lawyer therefore remains responsible for judgment even when AI performs much of the preliminary work.
This makes the “legal engineer” potentially more than a technical consultant. The role sits at the intersection of legal practice, software engineering and AI workflow design.
The partnership represents a way to scale specialized expertise without building every industry-specific implementation capability internally. The company launched its Partner Network in June, with a broader goal of enabling consultants and partners to deploy its technology inside organizations.
The legal industry is already crowded with specialized AI providers, including companies such as Harvey and other legal technology firms. Reuters reported earlier this year that AI companies were increasingly competing for law firms and even law students as the next generation of legal professionals becomes an important customer base.
The emerging competition is therefore not simply about whose model is most capable. It is increasingly about who can make AI reliable, usable and deeply embedded in professional environments. OpenAI’s partnership with Telon points toward that next phase.
The future of AI in law may depend less on replacing lawyers than on redesigning how lawyers work—with AI handling increasingly complex tasks while human professionals retain responsibility for interpretation, strategy and accountability.