The latest news from law and accounting tech shows a common pattern: firms want AI that improves delivery, but they also want control over proprietary knowledge, client data, and quality. That makes custom AI, workflow automation, and agentic systems a strategy issue, not just a software purchase.
AI is moving from tools to control points
The legal market news emphasizes AI sovereignty: firms are being pushed to think carefully about what they reveal to AI vendors in exchange for better performance. The concern is not only efficiency, but also how much proprietary knowledge leaves the firm as models improve.
For law and accounting leaders, that makes vendor selection and workflow design part of the same decision. The firms that win will likely be the ones that use AI to support internal expertise instead of outsourcing the firm's judgment to a generic model.
Accounting firms are being reshaped around who does the work
The accounting news points to a structural change in how firms operate. AI is taking over repetitive base-level labor, which changes the traditional staffing pyramid and pushes firms toward a more middle-heavy model focused on judgment, insight, and client value.
That shift creates room for custom workflows that route routine tasks to automation and reserve people for exceptions, review, and advisory work. In practice, that means firms should map which tasks are repeatable, which need supervision, and which should stay with senior professionals.
Risk and controls are a strong use case for agentic AI
Deloitte's new ControlCatalyst.AI launch shows where the market is heading: using generative AI and agentic AI together across internal audit, SOX, risk and controls, and compliance. The value proposition is not just speed, but smarter risk identification, deeper insights, and better resource allocation.
That is a useful pattern for smaller firms too. The best early workflows are often the ones with clear inputs, defined outputs, and strong business rules, such as compliance review, document triage, or issue detection.
Policy changes reinforce the need for adaptable client workflows
The Social Security proposal and the tax-cut campaign coverage are reminders that client-facing work is shaped by changing policy, even when no law has changed yet. Clients still ask questions before proposals become law, and firms need a way to respond quickly and consistently.
Custom AI can help by turning evolving policy into reusable client alerts, intake questions, and draft explanations. The goal is not to replace professional advice, but to make sure the firm can surface relevant changes faster and standardize the first pass of client communication.
Where firms should start
The practical takeaway from these stories is straightforward: AI should be built around the firm's workflows, not bolted onto them. Firms that define the work, the review points, and the data boundaries can get more value from automation without losing control.
That is especially important in professional services, where quality, confidentiality, and domain knowledge matter as much as speed. Custom AI and agentic workflows should support the firm's operating model, not quietly reshape it around a vendor's model.
- Treat AI sovereignty as an operating issue, not just a legal or IT issue.
- Start with repeatable workflows that have clear inputs, outputs, and review steps.
- Use AI to amplify mid-level judgment and client insight, not just reduce labor.
- Build client-facing workflows that can adapt as tax and benefits policy changes.
Sources watched
- Law Firms Need To Reassert Their AI Sovereignty, Here's How (Artificial Lawyer)
