Why your AI investments aren’t making your firm more valuable — and what to do about it

Over the past two years, law firms have made substantial investments in AI tools — everything from drafting assistants to research accelerators to meeting summarizers. And by most measures, these tools work. Lawyers are faster, more responsive, and more productive on a task-by-task basis. But the problem is, it’s not really showing up at the firm level. So why aren’t firms meaningfully more valuable as a result?

Andreessen Horowitz put it plainly: productive individuals don’t make productive firms. That dynamic is showing up clearly in AI adoption — 88% of organizations are using AI, according to McKinsey’s 2025 State of AI report, but only 6% are seeing meaningful bottom-line impact. The gap between individual productivity and firm-level value is real — and for law firms, it’s wider than most.

The problem isn’t the tools. It’s where they operate.

Most AI deployments today are horizontal. They improve how individual lawyers work, helping them draft faster or research more effectively. That’s useful, but it doesn’t change how the firm coordinates work, governs risk, or delivers value at scale.

Law firms don’t win because one partner is faster. They win because their collective judgement is sharper, their institutional knowledge is deeper, and their ability to execute across complex matters is more reliable than the competition. Individual AI tools don’t change any of that.

Meanwhile, the knowledge that actually differentiates a law firm — client relationships, matter history, deal precedents, pricing strategies — is scattered. It lives in email threads, in partners’ heads, in documents spread across systems that don’t talk to each other. A partner in London may not know a colleague in Dubai closed a similar deal last month. And when a senior partner leaves, their institutional knowledge often leaves with them.

AI can’t activate what it can’t reach — and in most firms, it can’t reach the knowledge that matters most.

The compliance problem is even harder

Generic AI tools face a second, more fundamental obstacle in legal: professional compliance.

Law firms operate under confidentiality obligations, ethical wall requirements, and regulatory frameworks that are both highly specific and professionally existential. A compliance failure isn’t just a bad outcome — it creates malpractice exposure and can threaten the firm’s ability to operate. That’s why, despite broad AI adoption, compliance concerns remain the number one blocker to firmwide AI deployment at law firms.

Generic AI wasn’t built for the constraints of legal work. You can’t feed client data into systems that don’t respect attorney-client privilege, ethical walls, or outside counsel guidelines. And because every firm’s compliance obligations are uniquely their own, there’s no off-the-shelf configuration that solves it.

The first wave of AI tools was built for a general environment. Law firms don’t operate in one.

What Firm AI actually looks like

The firms that are beginning to see real, durable value for AI are approaching it differently. Instead of asking which AI tool to buy, they start by asking what business outcomes they’re trying to achieve.

That shift in thinking tends to lead to a consistent set of choices:

• Start with the workflow, not the tool. Firm-level value comes from embedding AI into the processes that drive the business: business development, new matter intake, conflicts clearance, time capture, prebill review — not from adding AI to the periphery. 

• Put the firm’s own knowledge to work. When every firm has access to the same horizontal AI models trained on the same public data, differentiation comes from proprietary knowledge. Firms that can capture their client relationships, matter history, and internal methodology and use it to power AI gain a structural advantage that generic tools can’t replicate. 

• Build compliance in from day one. AI that operates inside a law firm must natively respect ethical walls, confidentiality obligations, and client billing requirements. These can’t be bolted on as an afterthought. 

• Measure qualitative improvement, not just efficiency. The most meaningful gains aren’t just about speed. They’re about making processes more robust, bringing more relevant information into decisions, and enabling firms to handle complexity that would be difficult to manage manually.

The next 12 to 18 months will define the next decade

The firms that move now — deliberately, with clear outcomes in mind — will build a competitive position that’s difficult to replicate. Not because they moved fast, but because proprietary knowledge, embedded workflows, and compliance-native AI compound over time. The gap between firms that get this right and those that don’t will widen quickly.

The good news: the infrastructure to do this well already exists. The question is whether your firm is using it.

To go deeper on what Firm AI looks like in practice across business development, profitability, and compliance, watch the on-demand recording of our recent webinar, Intelligence Applied: The next generation of AI in law. Our speakers, who bring more than a century of combined legal experience, walk through how agentic AI is changing what’s possible for firms like yours.

Watch the on-demand webinar: https://www.intapp.com/webinar/next-generation-ai-law/