The Defense Legal AI Gap: Why Plaintiff Firms Are Going Agentic While Defense Lags

Over the last two years, generative AI has moved from novelty to actual infrastructure (source). There are a plethora of AI options for the plaintiff bar, but very few, relatively, for the defense bar. This may sound peculiar, but just follow the money!

If one scans the legal AI landscape, you’ll notice something striking: plaintiff-side firms are flooded with purpose-built AI tools, from intake automation to demand generation platforms like Eve, while defense-only practices remain comparatively quiet territory.

That imbalance isn’t accidental.

Plaintiff firms run on contingency economics, high-volume workflows, and repeatable case patterns, making them ideal environments for automation that accelerates triage, document review, and settlement strategy. The ROI is direct and measurable: faster resolutions and stronger case selection.

Defense firms, meanwhile, operate in bespoke, risk-sensitive environments tied to billable hours and institutional client expectations, where efficiency gains can feel like revenue trade-offs. What defense firms need, in short, isn’t speed, per se, but really operational excellence from AI.

In the sections below:

Section 1 explains how litigation economics shaped today’s AI moment, then brings us to the present adoption reality.

Section 2 breaks down the three forces driving the defense lag (incentives, sales cycles, and where money flows).

Section 3 anchors it with current numbers.

Section 4 looks ahead to what changes in the next couple of years, and what defense firms can do now to turn AI from a threat into margin protection.

Feel free to also check out my blog article on Jeannie, CustomAI Studio’s defense-only legal AI solution here.

In this section, you’ll first get the historical “why” behind today’s split adoption patterns between plaintiffs and defense firms. Then we’ll zoom into the present: AI is widely used, but unevenly institutionalized, thereby creating a gap between individual experimentation and firm-grade, client-approved workflows, particularly on the defense legal side.

Section 2: Today, The three forces behind the defense lag

This section isolates the three drivers of the plaintiff/defense asymmetry.

First is incentives (what happens when efficiency threatens revenue vs. when efficiency creates revenue).

Second is the buyer journey (who must approve what, and why defense procurement is structurally slower).

Third is where money flows (plaintiff growth spend vs. defense cost-control spend), which quietly determines who builds the better AI stack.

Section 3: By the Numbers in legal defense AI

The data shows that AI adoption across the legal ecosystem is accelerating rapidly: 79% of legal professionals report using AI, in-house GenAI usage has surged from 23% to 52% in a year, and organization-grade adoption has nearly doubled, yet governance and value capture remain uneven, with 44% of firms lacking formal AI policies and nearly 60% of clients seeing no cost savings from firms still billing ~90% of work hourly.

At the same time, plaintiff-side firms are deploying capital aggressively: over $2.5B in advertising spend (with Morgan & Morgan alone at ~$218M), meaningful GenAI penetration in personal injury, and venture-backed platforms like Supio raising $60M to power plaintiff workflows. Defense-side firms, by contrast, operate within slow-moving procurement ecosystems (large vendor stacks averaging ~2,000 vendors and >$150M in spend) and suffer from revenue-ops friction (5 days to enter time, 47 to bill, 64 to collect).

The implication: plaintiffs are behaving like scaled, tech-enabled growth businesses, while defense firms remain process-heavy, hourly-driven, and operationally sluggish. If defense firms do not aggressively integrate AI into workflow, pricing, and operations, an I mean not just experimentation but structural adoption, they risk falling further behind faster, better-capitalized, AI-native plaintiff competitors.

Section 4: The next twenty-four months

The next wave is not “more ChatGPT.” It’s agentics: systems that monitor events, route tasks, draft outputs, update systems, and escalate exceptions, all under human supervision (or under “bounded autonomy”). 

In other words, AI shifting from answering to doing. This isn’t speculative since it’s already how the most operationally-minded firms are approaching automation … read more here.

Final Thoughts:

Plaintiff firms are moving faster on AI for the same reason they spend aggressively elsewhere: contingency economics reward throughput and speed, and the market is building plaintiff-specific AI products in response. Defense firms, in contrast, aren’t “behind,” but rather, they are constrained by incentives, governance, and buyer complexity.

But those constraints are exactly why defense has the biggest upside from auditable, human-in-the-loop agentic operations: fewer missed deadlines, tighter billing compliance, faster cycle time, and a clearer value story when pricing inevitably shifts away from pure hours.

It has come time for the defense to utilize the tech, but in different ways, that the opposing counsel is using daily.