The latest news across legal, accounting, and regulation points to a common theme: firms are moving beyond off-the-shelf AI and toward systems that reflect their own expertise, risk tolerance, and operating model. That shift matters for any professional-services firm thinking about where AI creates real leverage versus where it only adds another tool.
Big firms are treating AI as infrastructure, not a side tool
The most striking legal-tech headline is Kirkland & Ellis building a proprietary AI system for internal use. The stated point is not to chase the lowest common denominator, but to apply partner-level expertise across matters in a way that better reflects how the firm actually works.
For other firms, this is a useful signal: the firms likely to get the most value from AI are not just buying access to generic chat tools. They are designing internal systems that pull from their own knowledge base, fit their matter or engagement process, and support the way senior professionals review work.
Custom AI workflows create more value when they are tied to firm knowledge and review stan
A custom workflow is most useful when it helps staff move faster without weakening judgment. In practice, that can mean AI that drafts, summarizes, classifies, routes, or retrieves firm-specific information before a lawyer or accountant reviews the output.
This is where agentic workflows become more relevant than a simple chatbot. Rather than asking a model one question at a time, firms can build sequences that gather inputs, check them against internal guidance, and hand off work to the right person at the right stage. That is especially important in professional services, where quality control and confidentiality matter as much as speed.
Regulators are also moving toward system-level oversight
The PCAOB's move to create an Inspections Modernization Council suggests a broader shift toward evaluating firms' systems of quality control, not just isolated work products. That is relevant for accounting firm leaders deciding how to introduce automation without creating new oversight gaps.
The more AI becomes part of production work, the more firms will need controls around data sources, review paths, exception handling, and documentation. In other words, the same discipline that supports audit quality can also support safe AI adoption: define the process, monitor the process, and be able to explain how decisions were made.
Why this matters now for law and accounting firm leaders
The news cycle is telling firms two things at once. First, third-party AI tools are raising the baseline for speed and access to knowledge. Second, the firms that want lasting advantage are investing in systems that capture their own know-how and operational judgment.
That combination is a strong case for starting with a narrow, high-value workflow rather than trying to "AI-enable" the whole firm at once. Intake, research support, matter triage, document assembly, quality review, and internal knowledge retrieval are all stronger candidates than broad, open-ended chat use. The goal is not to replace professionals; it is to make the firm's expertise easier to deploy consistently.
- Treat AI as a firm workflow and governance project, not just a software purchase.
- Start with a narrow use case where your own knowledge and review standards clearly matter.
- Build controls around inputs, outputs, escalation, and documentation before expanding AI across the firm.
- Use generic tools where they help, but reserve custom workflows for work that reflects your firm's real value.
Sources watched
- Claude Opus 4.8 First Impressions (AI Daily Brief)
- PCAOB Establishes Inspections Modernization Council (CPA Practice Advisor AI)
