What tax professionals actually want from AI
Historically, tax research started with a search box. Professionals entered queries, reviewed documents individually and manually assembled answers using their own judgement.
Today, expectations are different. Tax professionals want contextual guidance before diving into complex research. They expect AI-driven tools that help frame issues, identify relevant authority and navigate efficiently — while still grounding conclusions in trusted content. Increasingly, they also expect conversational research that can interact with both vetted tax libraries and their own internal documents.
AI-driven tax research must reduce risk, not introduce it. Platforms that rely solely on primary source material leave interpretation and judgement entirely to the model, increasing the likelihood of unsupported or hallucinated conclusions that cannot withstand audit or regulatory scrutiny. The real value of AI emerges when expert-verified secondary content and editorial judgement are combined with primary authority, producing conclusions that are properly contextualized, validated and complete.
This evolution only works inside controlled environments. When authoritative research content and firm knowledge are analyzed together within closed systems, firms gain efficiency without sacrificing governance. That distinction separates enterprise-grade AI from ad hoc or consumer-orientated tools that lack accountability.
AI adoption reflects this shift. Seventy percent of U.S. firms now use AI at least weekly, with advanced use embedded in tax and audit research workflows. The most successful implementations operate inside integrated, controlled environments where authoritative content and internal knowledge coexist.
Platforms such as CCH® AnswerConnect exemplify this approach by layering large language models on top of curated content, continuously maintained content. Editorial teams monitor federal and state law changes, apply expert analysis and connect updates across a broader network of authority — ensuring AI outputs deliver productivity gains while remaining aligned with structured tax law, not open-internet interpretation.
From questions to clarity: How modern tax research simplifies complex discussions
Customer expectations are raising the bar
Large firms are increasingly explicit about what they expect from AI-powered research tools. They want automation and generative capabilities, but only when answers are trusted, sourced and clearly cited. They also expect to deploy these tools across a broad employee population with confidence in both content quality and security posture.
Equally important, firms expect that their proprietary data will not be used to train someone else’s model.
These expectations are shaping purchasing decisions as firms scale AI adoption. Confidence in content integrity, governance models and data protections has become decisive.
What 'great' looks like
The firms seeing the most successful AI adoption are deliberate. They’re not chasing every new feature. Instead, they prioritize governance, security and content integrity early — recognizing that trust is what ultimately enables scale.
Moving slightly slower at the outset allows firms to move faster over time. When controls are built in from the beginning, professionals trust the system, adoption increases and AI can be deployed confidently across the organization.
AI undeniably improves speed and efficiency in tax research. But speed alone is not enough. Accuracy, transparency and protection of client data remain non-negotiable. The firms truly ahead understand that balance — and are choosing AI systems grounded in authoritative content vetted by both primary and secondary sources and designed for the realities of regulated, high-stakes work.
This article first appeared in Forbes.
