.jpg)
One industry has gone from barely experimenting with AI to widespread adoption in just two years. AI use in the legal field jumped from 23% to 78%, outpacing adoption in both finance and healthcare.
According to Litify’s third annual State of AI in Legal Report, which surveyed hundreds of professionals across law firms, corporate legal departments, and plaintiff practices, legal professionals are now among the fastest adopters of AI anywhere.
But the headline number hides a deeper issue.
Only 14% of respondents say AI is helping them reduce costs, and just 7% report billing more time. In other words, many firms rushed to acquire powerful AI tools but are barely using their full capabilities. The gap between using AI and improving the economics of legal work remains large—and continues to grow.
Curtis Brewer, CEO of Litify, describes this as an “AI maturity gap.”
“A firm that relies solely on a general-purpose tool like ChatGPT is only at the first step of its maturity journey.”
Why General-Purpose AI Hits a Ceiling
Litify’s data shows that most firms are relying on broad AI tools rather than platforms designed specifically for legal work.
ChatGPT: 66% usage
Microsoft Copilot: 42%
Google Gemini: 24%
These tools are being used primarily for legal research (66%) and document summarization (39%). However, only 6% use AI to generate invoices, and 5% use it for client communication.
In other words, firms are applying AI to tasks that feel productive but rarely impact revenue or operational efficiency.
The issue isn’t that general AI tools are ineffective—they simply reach a plateau.
That’s why legal-specific AI platforms are gaining traction. Companies like Harvey, trained on case law, contracts, and regulatory frameworks, are designed to understand the nuances of legal work.
Harvey now counts PwC, A&O Shearman, and roughly half of the 100 highest-grossing law firms in the U.S. as clients. The company has raised more than $1.2 billion and is reportedly seeking another $200 million round at an $11 billion valuation.
Brewer explains the difference:
“Legal work is defined by nuance—jurisdictional requirements, compliance rules, solicitation standards, and practice-area-specific workflows. General models often overlook those details.”
Another limitation is context. A general AI model only knows what you include in a prompt. It cannot access the broader history of a case, a client’s background, or the firm’s internal workflows. And after producing an answer, it cannot take further action.
Legal-specific tools, by contrast, can operate inside a firm’s systems—summarizing a case while also suggesting next steps, identifying missing information, or recommending additional questions.
The Shadow IT Risk
The rapid rise of AI adoption has also created a governance problem.
While 78% of legal professionals say they use AI, only:
41% of firms have a formal AI policy
45% provide adequate staff training
That leaves a significant portion of employees experimenting with AI tools in what amounts to shadow IT—technology used without formal approval or oversight.
Brewer stresses the seriousness of the issue:
“Security, security, security! Given the highly sensitive nature of legal data, leaders should be concerned that nearly a third of staff may be using AI without IT oversight.”
The risk is obvious. Lawyers or staff may paste confidential client information, sensitive case details, or even HIPAA-protected medical records into public AI systems. One careless prompt could trigger a data breach, regulatory violation, or the loss of client trust.
Brewer argues the root problem is organizational:
“If firms fail to provide clear guidance and purpose-built tools, employees will find their own solutions.”
Without structured adoption, companies risk losing efficiency gains while exposing themselves to new security vulnerabilities.
From AI Assistant to Business Driver
The real value of AI comes from workflow integration.
Take billing as an example.
Using ChatGPT to generate an invoice template is like using your phone’s calculator instead of a full accounting system. It works—but you still have to manually enter every client detail, payment amount, and line item. You save a few minutes creating the template, but spend much longer filling in the rest.
Integrated AI systems work differently.
When AI is connected directly to case management, billing, and client records, it can:
Automatically generate invoices with pre-filled client details
Suggest missing billable time entries
Detect billing inconsistencies or errors
Flag potential write-offs before they occur
According to Brewer:
“When AI lives alongside your billing, client, and case workflows, it transforms from an assistant into a proactive business partner.”
Some firms using this level of integration have dramatically expanded their capacity. Litify reports that certain clients can now handle twice as many cases with the same staff, while the fastest-growing firms have expanded headcount by up to 400% as they scale into regional and national markets.
Closing the AI Maturity Gap
Brewer suggests firms must evolve across four dimensions to unlock AI’s full value.
1. Tools
Relying solely on ChatGPT is not enough. Firms need legal-specific platforms that integrate with case management, billing, and client systems.
2. Readiness
Every firm should establish a formal AI policy defining approved tools, data-handling practices, human review requirements, and incident response procedures. Training must be treated as essential—not optional.
3. Task Scope
Many firms stop at research and summarization. The next stage is workflow automation: routing requests, performing conflict checks, building case chronologies, and automating intake. Eventually, AI can help assign cases, generate invoices, and manage client onboarding.
4. Impact Measurement
Before investing further, firms should define metrics such as:
Cost per matter
Case turnaround time
Billing write-off rates
Error rates
Brewer believes the experimental phase is ending.
“The try-it-and-see period is over. Leaders will expect clear ROI.”
The Real Competitive Divide
The firms pulling ahead are not just purchasing AI tools—they are redesigning how legal work happens, from intake to invoicing and from research to billing.
Training, governance, and measurable performance improvements are built into the process from the start.
The alternative is continuing to operate powerful technology at minimal capacity.
You can drive a sports car in first gear—but eventually, a competitor will figure out where the other gears are.