US legal tech has moved from experiments to essentials. Firms are shifting from ad-hoc prompting to auditable AI-enabled workflows and agentic adoption, while clients increasingly expect “more for less.” Smaller, agile practices appear to be outpacing BigLaw on real implementations.

Why it matters

• Trust and transparency now determine adoption. Auditable systems, clear governance, and cultural buy-in are the differentiators—not flashy models.
• The capability gap is narrowing. AI compresses build times, letting new vendors emerge quickly—great for choice, risky for reliability. Expect consolidation.
• Client pressure is peaking. Users know AI accelerates work and want pricing to reflect productivity gains (e.g., the provocative notion of an “AI hour”).

What we heard (themes & insights)

1) Adoption: from pilots to production

• Use has surged across firms; the early “try anything” phase has given way to governed, workflow-led deployment. Lindsey notes every LLM can hallucinate—so governance at personal and organisational levels is non-negotiable to ensure accuracy.
• Gannon referenced the dramatic 354% year-on-year growth in AI use among law-firm professionals, with further acceleration expected.

2) From prompting to workflows

• Training proliferates, but the centre of gravity is shifting from “prompt skills” to productised workflows and agents that reduce reliance on free-form prompting.
• Auditable, compliant platforms (including on-prem/private cloud options) are building trust by eliminating the “black box.”

3) Governance, ethics, and culture

• Rapid capability gains create regulatory and ethical pressure. Clear internal policies and transparent client messaging are uneven across firms; some lead with openness while others stay vague.
• Change is people-first: anxiety about role change and billing models can stall programmes. Leaders need both IQ and EQ to land the shift.

4) Market dynamics: mushrooms, then mergers

• The barrier to building has collapsed; “mushroom” vendors are everywhere. Selection should prioritise white-glove implementation, support capacity, upgrade cadence, and fit to bottlenecks over shiny demos.
• With long sales cycles and service constraints, consolidation is likely. Quarterly release cycles are already too slow for AI’s pace.

5) Small beats big (for now)

• Smaller firms are winning through targeted use-cases (e.g., document review, due diligence, clause extraction), building momentum one workflow at a time—and expanding from proven value. Larger firms often risk “innovation theatre.”

6) Clients, pricing, and the “AI hour”

• Clients use AI personally and expect efficiency dividends. The proposed “AI hour”—pricing time saved differently—signals a rethink of value, transparency, and collaboration (including when clients start with AI-generated drafts).

7) Access to justice

• AI should expand access for those priced out of legal help—the challenge: balancing quality, independence, and regulation while broadening service reach.

Playbook: what firms can do next

1. Pick a bottleneck, not a platform. Start with a single high-friction workflow; ship value fast, then scale sideways.
2. Mandate auditability. Require extraction of inputs/outputs, traceability, and policy hooks. Consider on-prem/private cloud where needed.
3. Shift training to roles and runbooks. Teach teams to operate workflow tools and review AI outputs, not just to prompt.
4. Price transparently. Pilot AI-aligned pricing (including variants of the “AI hour”) with clear value narratives.
5. Resource implementation. Select vendors for service depth and upgrade velocity; check referenceability and support ratios.
6. Lead with empathy. Communicate role evolution, not replacement; reward adoption; give space for learning.

The takeaway

The winners will be the firms that ship audited workflows, measure value, price fairly, and bring their people with them. The technology is ready enough; culture, clarity and client-centred design will decide the gap between theatre and transformation.

But are the people ready?