AIToday
Large Language ModelsAI Safety & AlignmentAI Business & IndustryYahoo Finance AIPublished: Aug 16, 2026, 06:00 JST5 min read

Intapp's AI Honesty Problem: Law Firms Need Permission, Not Just Answers

Intapp's AI Honesty Problem: Law Firms Need Permission, Not Just Answers

Key takeaway

  • Intapp, a software vendor serving law firms and professional partnerships, launched Celeste, an AI platform that enforces access controls and ethical walls—preventing employees from accessing confidential information they are not permitted to see.

  • Unlike ChatGPT and similar horizontal AI assistants, which answer questions truthfully regardless of who asks, Celeste knows each user's role and permissions, matching the compliance rules that apply to human employees.

  • The company's 25 years of law-firm-specific data infrastructure and compliance expertise give it a rare competitive advantage in solving what it sees as the real AI problem in professional services: not hallucination, but honesty without permission.

3 Key Points

  1. What happened

    Intapp, a software company serving law firms and similar partnerships, launched Celeste, an AI platform designed to enforce access controls—preventing employees from seeing documents or information they are ethically walled off from. Unlike generic AI assistants, Celeste knows who is asking and what they are permitted to know, leaving an audit trail for regulators.

  2. Why it matters

    The core risk at law firms is not AI hallucination but the opposite: a generic AI assistant answering truthfully about deals, clients, or confidential information that an employee has no right to access. Intapp's 25 years of firm-specific data and compliance infrastructure give it a rare ability to solve this problem at scale—a competitive moat that horizontal AI vendors cannot easily replicate.

  3. What to watch

    Celeste reached general availability on July 15, after Intapp's fiscal year ended June 30. The company reported cloud annual recurring revenue of $495.7 million (up 29% year over year) and adjusted net income of $103.6 million, but the real test is how quickly Celeste gains traction and generates revenue—a capability that required two decades to build.

In Depth

Read the full story

Intapp was founded in response to a complaint, not a business plan. In the early 2000s, a Silicon Valley data integration startup applied for a patent. The attorney handling the paperwork recognized that his own law firm was a disaster in exactly the domain the patent addressed and asked if he could license the product. When Intapp agreed, that lawyer—who was the CIO at his firm—quit his job, joined Intapp as its first salesperson, and introduced the company to every law firm he knew. Referrals sustained the company for two decades without venture capital; its first outside money arrived with the 2021 IPO.

Intapp's customers are what CEO John Hall calls "large partnership firms": law firms, accounting firms, investment banks, private equity shops, and consultancies. Hall estimates this category generates around $4 trillion in annual revenue in the United States. Building software for these firms is tricky because it is not enough to build a functional product; the software must understand which of a firm's thousands of employees may see which document and be able to prove it in an audit. Generic AI vendors lack the lived-in data set required to copy Intapp's approach, which is why most software companies simply do not chase this market.

In February, Intapp launched Celeste, an agentic AI platform (a system that can take actions, not merely answer questions). It reached general availability on July 15. Hall distinguished Celeste from other legal AI tools—such as Harvey, an Intapp partner, and Legora—as well as from horizontal assistants like ChatGPT Enterprise. "One of the problems with the horizontal systems is that they tell the truth," Hall said. "But you may not be a person who is supposed to get that answer inside the firm." The failure mode of a chatbot at a law firm is not hallucination; it is honesty. Someone asks about a deal they are ethically walled off from, and a regular AI assistant helpfully tells them about it anyway. Celeste is designed to know who is asking. Ethical walls, material nonpublic information rules, and independence requirements apply to the software the same way they apply to employees. A partner walled off from a deal gets no answer from Celeste. Every interaction leaves an audit trail, because regulators may have questions about it.

Intapp's fiscal year closed on June 30 with strong results. Cloud annual recurring revenue (ARR) reached $495.7 million, up 29% year over year. Total ARR was $590.5 million, up 22%. Top-line revenue came in at $577.8 million for the year, a 15% increase. The company is profitable on an adjusted basis: adjusted net income was $103.6 million, and free cash flow hit $144.7 million, or 25% of revenue—three years ahead of management's long-term targets. The number of clients paying more than $1 million a year grew from 109 to 142. Microsoft co-sold eight of the 10 biggest deals. More than 30 cloud migrations were signed in the fourth quarter, a company record. As of August 14, shares are up roughly 20% since the earnings report and 93% over the past three months, though they remain down 16% from December's all-time peak.

At a $3.07 billion market cap, Intapp trades around 5.3 times sales and 20.4 times forward earnings estimates. The average analyst price target of $39.43 sits below the current quote. Management sees a $50 billion addressable market for agentic professional services software. Since Celeste only became generally available after the 2026 fiscal year ended, its market traction and monetization are what investors will watch—a challenge with no shortcuts to 25 years of knowing exactly how a law firm thinks.

Context & Analysis

Intapp's path to dominance in professional services software is unusual for Silicon Valley. The company was never venture-backed; its first outside capital came with a 2021 IPO. Instead, it grew through direct referrals from a single satisfied customer—the law firm that first licensed its data integration product in the early 2000s. That two-decade focus on one narrow market created what CEO John Hall calls the company's competitive moat: accumulated firm data and compliance infrastructure that cannot be easily copied by large, multipurpose AI vendors. The stakes are clear: building a large language model is one problem; building one that knows which of a firm's thousands of employees may not access a given fact, and proving it in an audit two years later, is a different and harder problem entirely.

The arrival of Celeste, an agentic AI platform launched in February and reaching general availability on July 15, represents Intapp's answer to what it frames as the real AI risk in law firms: not hallucination, but honesty without permission. A generic AI assistant like ChatGPT Enterprise will truthfully answer questions about deals, clients, and confidential information that an employee is ethically walled off from—often with minimal prodding. Celeste enforces the same access rules and ethical walls that apply to human employees, leaving an audit trail that satisfies regulatory scrutiny. This is not a feature that can be bolted onto a horizontal AI platform; it requires the kind of lived-in data set and compliance logic that Intapp has spent 25 years building.

FAQ

When did Celeste become available?
Celeste was launched at a product event in February and reached general availability on July 15.
What is the core difference between Celeste and tools like ChatGPT Enterprise?
A generic AI assistant answers questions truthfully to the best of its ability, regardless of who is asking. Celeste is designed to know who is asking and whether that person is permitted to see the answer—enforcing the same ethical walls and access rules that apply to employees, with an audit trail for regulators.
How large is the market Intapp targets?
Management estimates that large partnership firms (law firms, accounting firms, investment banks, private equity shops, and consultancies) generate around $4 trillion in annual revenue in the United States. Management sees a $50 billion addressable market specifically for agentic professional services software.
Yahoo Finance AIRead Original Article

Get the latest Large Language Models news every morning

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · takes 30 seconds · unsubscribe anytime

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next article10 AI Video Tools for 2026: Topview Leads Workflow-First Automation

The AI news that matters, in one minute each morning.

Sign up free