Back to news
NewsJune 25, 2026· 3 min read

Shoosmiths deploys AI contract reviewer trained on its own deal expertise

UK law firm Shoosmiths launched Project Apollo, a generative AI tool built with Microsoft that explains contract changes by referencing the firm's playbooks. Designed to teach junior lawyers why changes matter, not just what changed.

Our Take

Explainability wrapped as junior-lawyer training is a credible product angle, but the firm has not published independent benchmarks, deployment numbers, or learning outcomes to prove the tool actually accelerates skill development.

Why it matters

Law firms are building proprietary AI tools to embed institutional knowledge and competitive advantage into junior hiring pipelines. Shoosmiths' bet on transparency and learning signals that firms see AI adoption as a retention and upskilling lever, not just a cost-cut.

Do this week

Legal tech leaders: before greenlighting an internal contract AI, audit whether your firm can define and measure what 'faster learning' actually looks like (accuracy gains by junior associate, reduced partner review cycles, months-to-competence reduction) and set that as a success metric now.

Shoosmiths built and deployed a contract review AI trained on its own guidance

Shoosmiths, a UK law firm, unveiled Project Apollo on 24 June, a proprietary generative AI tool designed to review and mark up contracts. The tool was built over a year-long development and pilot cycle in partnership with Microsoft and runs in Microsoft's Azure environment.

Unlike commercial contract-review AI platforms that rely on general language models, Project Apollo is trained against Shoosmiths' own playbooks, gold-standard drafting, and institutional knowledge. The firm designed the tool to surface its reasoning at every step, explaining not just what changes it recommends but why, with references back to the underlying firm guidance and dealmaking precedent.

The tool has already been deployed across Shoosmiths' internal operations. David Jackson, chief executive, said the platform allows the firm to "deploy its collective dealmaking expertise at scale, allowing lawyers to not only see what amendments have been made, but most significantly, why."

Explainability is being packaged as a training and retention tool

Shoosmiths frames Project Apollo as a learning instrument, not merely a productivity gain. The firm argues that junior lawyers benefit from seeing senior-level reasoning in real time, with explanations curated to accelerate skill development. In a sector where knowledge transfer depends on close mentorship and informal exposure to deal patterns, embedding firm expertise into an AI interface is a plausible way to scale that exposure.

The emphasis on transparency also sidesteps a real friction point for law firm adoption of AI: many junior and mid-level lawyers distrust black-box contract suggestions because they cannot verify reasoning or defend a recommendation to a partner. Shoosmiths is betting that explainability removes that friction and makes the tool feel like a learning partner rather than an automation threat.

The partnership with Microsoft is notable chiefly for its infrastructure and credibility. Azure hosting and Microsoft endorsement signal enterprise-grade deployment, but the real IP is the firm's own knowledge base and the decision to surface it.

The test is learning outcomes, not feature announcements

Shoosmiths has not published benchmarks, deployment metrics, or learning-acceleration data. There is no independent verification of whether junior lawyers actually spend less time reaching competency, whether partner review cycles shrink, or whether the tool's explanations materially improve decision-making quality. The firm's CEO quotes emphasise culture and knowledge sharing, but culture is not proof of impact.

For other law firms considering similar builds, the question is not whether explainability is a good design choice (it is), but whether Shoosmiths can show measurable change in junior lawyer velocity, retention, or partner confidence in early-career output. Until that data surfaces, Project Apollo is a credible product story, not a proven one.

The firm will publish further commentary on its tech strategy in coming days, per the reporting. Watch for any disclosure on adoption rates, performance metrics, or learning-outcome measurement.

#Legal AI#Enterprise AI#LLM
Share:
Keep reading

Related stories