What is Intapp Walls for AI and how does it work with Harvey?

Intapp Walls for AI is a technology integration that brings Intapp's ethical wall enforcement — previously a standalone compliance product — directly into Harvey's AI deployment layer. When an attorney using Harvey queries documents or asks the AI to analyze a matter, Intapp Walls for AI ensures the AI cannot surface information from matters that are on the other side of an ethical wall. The wall enforcement happens at the AI query level, not just at the document retrieval level.

Why does AI conflict of interest matter for law firms using Harvey?

When you use AI to analyze client documents, the AI model sees everything in scope unless barriers are explicitly enforced. Without ethical wall controls, there is a risk that querying Matter A could surface privileged information from Matter B if both clients' documents are accessible. This is especially acute at multi-client firms handling matters across competing parties in the same industry. Ethical wall violations — even accidental ones caused by AI — carry the same professional responsibility consequences as manual violations.

Is this integration available for small law firms?

The Intapp + Harvey partnership was announced in February 2026 and is available to firms of varying sizes that use both Harvey and Intapp products. Small firms with fewer conflicts scenarios have less immediate urgency, but the integration is relevant to any firm that manages concurrent representation, has practice groups that could face conflicts, or is being asked by enterprise clients about AI governance protocols.

What was the previous objection to deploying Harvey firm-wide at multi-client firms?

The core objection was that AI queries could theoretically surface documents from conflicting client matters without a reliable enforcement mechanism. Even if attorneys were disciplined about what they asked, the AI model's training context or document scope could create exposure. Ethical wall enforcement had to be manually managed — which is difficult at scale. The Intapp integration automates this enforcement at the deployment level, making it reliable rather than dependent on individual attorney behavior.

How should a small law firm think about AI governance before deploying tools like Harvey?

Start with three questions: (1) Which matters have existing conflict checks active, and can your AI deployment respect those same barriers? (2) Are client documents stored in a way that allows per-matter access controls — or is everything in a shared document library? (3) Do your engagement letters disclose that AI tools will be used to assist on the matter? Governance infrastructure built before deployment is far easier than retrofitting it after attorneys are already relying on AI outputs.

AI Conflict of Interest Multi-Client Firms

Published: June 2026 | By: The Crossing Report

There is one objection that consistently stops law firm AI deployments in their tracks.

Not cost. Not training. Not vendor lock-in. It is this: what happens when the AI sees something it should not?

In a multi-client firm, that question has teeth. An attorney uses Harvey to analyze a contract. The AI is helpful. But what if the AI's access scope includes documents from a client on the other side of an ethical wall — another matter, another partner, another practice group the firm has walled off from this work? The wall exists on paper. Does it exist in the AI?

Until recently, the honest answer was: it depends on how carefully you configure things manually. Which means it depends on human discipline. Which means it is not actually enforced.

In February 2026, that changed.

What Intapp and Harvey Built

On February 23, 2026, Intapp and Harvey announced a strategic partnership to bring ethical wall enforcement directly into Harvey's AI deployment layer.

The product is called Intapp Walls for AI.

When an attorney using Harvey queries documents, asks the AI to analyze a matter, or runs any AI workflow that involves firm data, Intapp Walls for AI enforces the firm's existing ethical wall framework at the query level. The AI cannot surface information from matters that are on the other side of a wall.

This is not a new ethical wall product. Intapp has been the dominant vendor in legal conflicts and information barrier technology for years. What is new is that the wall enforcement happens inside the AI platform, not just in the document management layer.

Before this integration, a firm using Harvey could configure access controls at the document storage level. But if those controls were misconfigured, or if an attorney queried with broader context than intended, the wall enforcement and the AI deployment were separate systems that had to be manually coordinated.

With Intapp Walls for AI, they run as one system.

Why This Matters More Than It Sounds

Law firms think about ethical walls differently depending on their size and practice mix.

At an Am Law 100 firm representing both sides of an industry, ethical walls are a daily operational reality — active conflicts technology, information barrier notices, walled-off teams. The Intapp + Harvey integration is obviously relevant: you cannot deploy firm-wide AI without it.

At a 25-lawyer firm, the scenario feels more remote. Conflict checks are more manual. Practice groups do not formally wall off from each other. You might think this is not your problem yet.

But the pressure is moving down.

Enterprise clients are increasingly asking their outside counsel — including smaller and mid-size firms — about AI governance protocols before engaging. They want to know where firm AI systems can see their documents, who controls those permissions, and what prevents their confidential information from being accessible during work on a competing matter.

If your answer is "we trust our attorneys to be careful," that is a different answer than "our AI deployment enforces the same information barriers as our conflicts system." The second answer is now achievable. It was not a year ago.

The Governance Gap This Closes

The governance data from 2026 is consistent: firms are adopting AI faster than they are building governance frameworks around it.

Harvey's own firm governance study found that 69% of firms had no formal AI governance policy as of mid-2026. Karbon's 2026 benchmark found that 21% of accounting firms had an AI governance policy — roughly consistent with the legal sector. Deloitte's State of AI 2026 report found only 1 in 5 professional services firms with mature AI governance.

The ethical wall objection to AI deployment is one specific instance of this governance gap: firms are using AI on client work without mechanisms to enforce the same information barriers they apply to human attorneys.

Intapp Walls for AI closes that specific gap at the AI layer, not just the document layer.

That is the difference between a governance policy that says "attorneys must be careful about which client documents they upload to AI" and a governance control that enforces that care automatically.

What Smaller Firms Should Watch

The Intapp + Harvey partnership targets enterprise-grade deployments. If you are a 15-person firm on a basic Harvey subscription, you are not immediately configuring Intapp Walls for AI.

But the pattern matters for smaller firms for two reasons.

The architecture is the direction of travel. Legal AI platforms are being retrofitted with the same compliance infrastructure that law firms have always required. If you are evaluating an AI platform today, ask the vendor how ethical wall enforcement works at the AI query level — not just at document storage. If they do not have an answer, that is a gap in their product roadmap and a risk in your deployment.

Client-side pressure does not wait for your firm to grow into enterprise tier. A corporate client already asking about AI governance from their Am Law 100 counsel will eventually ask the same questions of your 30-person firm. The right time to build a governance answer is before you need it in a pitch.

One Thing to Do This Week

Pull up your firm's conflict check process. Ask one question: when your attorneys use AI tools to work on a matter, is there any mechanism that prevents the AI from accessing documents or knowledge from a conflicting matter — or is that solely managed by attorney discipline?

If the answer is "attorney discipline," document it as a governance gap. You do not need to solve it today. But knowing where the gap exists is the starting point for a governance policy that will hold up when a client or a bar auditor asks.

The tools to close this gap exist. Knowing what is possible changes what you can promise.

The Crossing Report covers AI adoption for professional services firm owners. For law firm AI governance updates, subscribe at crossing.one.

Harvey Now Enforces Ethical Walls Inside AI: What It Means for Multi-Client Law Firms