OpenAI's announcement that it is building dedicated AI tools for banking, investment and legal work—to be embedded in ChatGPT—is significant because it confirms what we already knew: the major AI labs are in a straight fight for the professional services market. Bloomberg reports that Codex agents will handle equity analysis, sales support and eventually legal work. But here is what matters to UK mid-market firms: this is capability racing, not governance racing. OpenAI is building features. Your regulator—whether that is the FCA under PS22/9, the SRA under the Code of Conduct, the FRC under ISA UK, or the PRA under SS1/23—is asking you to build controls. These are not the same thing.
The pattern is clear now. Every major AI vendor is rushing domain-specific tools. Harvey focuses on legal document work. Luminance sells compliance. Microsoft's Copilot is being bolted into Office workflows. Anthropic is in the same race as OpenAI. But what we are seeing is a homogenisation of capability without a corresponding answer to the governance question that keeps your compliance team awake. The EU AI Act is live in parts of the bloc already. The ICO is watching UK GDPR compliance in AI systems. Lloyd's Blueprint Two is asking the insurance market hard questions about third-party model risk. These tools are coming fast, but the frameworks for deploying them responsibly are not keeping pace. That is the real story.
Trovix's view is unfashionable but necessary: a general-purpose model with good governance infrastructure beats a specialized model with weak controls. ChatGPT with Codex agents is powerful. But if you cannot audit the decision, cannot explain the output to your clients or your regulator, and cannot segregate data properly, you have built a liability, not a tool. Some vendors sell the dream of plug-and-play specialisation. We sell something more boring: the ability to see what your AI actually did, why it did it, and whether it breached your obligations. That matters more than the vendor's feature roadmap. When Harvey or Legora or OpenAI's new legal tools make a call that affects a client outcome, you need to know it happened and be ready to defend it. Governance dashboards and audit trails are not marketing-friendly. They are mandatory.
What you should do now: do not wait for OpenAI's legal and finance tools to mature before you move. Begin with a governance-first AI adoption strategy. If you are already using LLMs—and most mid-market firms are—map what you are using, where data is flowing, and what controls exist. Use Trovix Audit to establish baseline governance. For intake, due diligence or regulatory change monitoring, build on tools that were designed with compliance in mind from the ground up. The firms that will win the next two years are not those that adopt the cleverest AI fastest. They are those that adopt it responsibly first. That is boring. That is also the difference between getting ahead and getting fined.
Source: Bloomberg News