OpenAI Plans 'Codex for Legal' Vertical as Big Tech's Race to Own the Legal Workflow Reaches Three-Way Competition with Anthropic and Microsoft
OpenAI is planning to launch a dedicated legal AI product, provisionally branded 'Codex for Legal', according to sources speaking to industry publication *Artificial Lawyer*. The offering would sit within a broader family of vertical enterprise tools under the Codex platform — OpenAI's facility originally built for software engineers, now being extended across major business sectors including finance and legal. The plan involves recruiting from the legal technology sector: an executive from a well-known contracts-focused legal tech company has already been approached, with additional senior hires under active consideration. The move directly parallels Anthropic's earlier expansion of Claude for Legal — which now encompasses 12 plugins and a growing range of MCP (multi-channel protocol, a standard that allows AI tools to connect to third-party software) connectors — and Microsoft's Legal Agent product. All three Big Tech operators are competing to embed their AI tools at the centre of lawyers' daily workflow, whether through word processors, document management systems, or standalone platforms. Freshfields and other major firms have adopted Anthropic's Claude for Legal. OpenAI's entry is being pursued alongside the rollout of its 'OpenAI Deployment Company' — a team of forward-deployed engineers sent into enterprise clients to help them integrate AI at scale. Greg Brockman, OpenAI's co-founder, has announced plans to integrate ChatGPT and Codex into a single system as part of the company's focus on agentic AI (AI that can take sequences of actions autonomously, rather than simply responding to individual prompts).
Why this matters
OpenAI entering the legal vertical with Codex for Legal creates a three-way Big Tech competition — alongside Anthropic and Microsoft — for control of core legal workflows such as contract review, document drafting, and discovery. For law firms, the arrival of a third major platform with embedded LLM (large language model — the AI technology underlying tools like ChatGPT) capabilities both broadens choice and intensifies the strategic decision of which platform to standardise on. Firms that go 'all in' on one provider risk locking themselves into a single LLM supplier as the technology evolves; specialist legal tech companies argue they can offer LLM-agnostic solutions that retain flexibility. The critical commercial implication is that commoditised legal tasks — basic document review, standard contract markup — face genuine automation pressure from these tools, reshaping how firms price and staff routine work. For City trainees, this signals that high-value judgment-intensive work will remain core, while volume-based tasks compress.
On the Ground
A trainee working on the legal technology side of a firm's AI adoption programme would review technology licence agreements and data processing agreements for any Codex for Legal or Claude for Legal deployment, draft AI governance policy provisions covering data security and model outputs, and prepare a vendor due diligence questionnaire comparing the Big Tech offerings against specialist legal tech providers.
Interview prep
Question you might get
“If a Magic Circle firm asked you to advise on the key contractual risks in entering a multi-year licence for a Big Tech legal AI platform, what provisions would you focus on?”
Sign up free to see the full answer
A model answer you can lift into an interview — how to frame this story for a partner.
Sign up freeSources
My notes
saved