How Generative AI Is Transforming Legal Practice With 70% Faster Summaries and 33% Faster Complex Drafting

How AI Is Saving Time and Changing Legal Practice at Shearman & Sterling

Shearman & Sterling is pushing the boundaries of AI integration in legal work. Since adopting generative AI tools like Harvey in 2022, the firm has developed its own AI system, Contract Matrix, in partnership with Microsoft. This system grounds AI in high-quality legal data, producing more accurate and relevant outputs.

From Early Experimentation to Client-Focused Innovation

Over the past three years, the firm has moved beyond basic AI experiments. David Wakeling, global head of the firm’s AI advisory practice, explains that AI adoption is no longer just internal efficiency—it has become a client service differentiator. Clients now approach the firm seeking advice on deploying AI responsibly within their own organisations.

Catherine di Lorenzo, partner in data protection and technology, highlights how the firm is creating tools that empower clients to self-assess AI systems against regulatory frameworks. This approach reflects a shift from traditional legal service delivery toward more accessible, tech-enabled solutions.

Making the Decision to Embed AI

The firm’s AI journey began with a clear focus on safety and governance. Before full deployment, they established strict data security protocols, including encryption and no data retention policies. Initial rollout was cautious—limited to certain teams and excluding sensitive areas like HR—to control risk and build confidence.

User feedback guided improvements, such as adding source citations to AI-generated content, which helped build trust in outputs. This step-by-step rollout minimized risks and maximized adoption.

Scaling AI Across Jurisdictions

Shearman & Sterling operates in 43 jurisdictions. To address differing local laws, each office builds its own legal data libraries within Contract Matrix. This ensures AI outputs reflect regional legal language and precedents.

Compliance is coordinated centrally, with local offices providing tailored legal advice. This model balances consistency with local expertise—a practical approach for global firms.

AI in Daily Legal Tasks

AI now handles far more than basic tasks like text rewriting or summarising. Lawyers can input lengthy documents and receive detailed answers grounded in specific legislation, such as the EU AI Act. Contract Matrix offers access to contract examples tailored by sector and jurisdiction.

Ensuring AI Output Is Trustworthy

To reduce AI hallucinations, the firm focuses on:

• Grounding AI in curated, high-quality data
• Using advanced foundation models
• Allowing lawyers to view source materials alongside AI-generated content

These measures ensure transparency and maintain lawyer oversight. Some firms even use two AI systems to cross-check outputs, though this doubles resource requirements.

Significant Time Savings

Independent client surveys show that Contract Matrix delivers up to 70% time savings on summarisation and about 33% on complex legal drafting. These figures highlight AI’s practical impact on legal workflows.

Ethics and Human Oversight

Human validation remains central. AI assists with repetitive, standardised tasks but does not replace lawyer judgment. Every AI result is reviewed, with lawyers confirming the facts and reasoning before use. This preserves accountability and quality, especially in litigation and bespoke cases.

AI Applications Beyond Contract Review

AI is also deployed in large-scale due diligence, M&A, and litigation support. For example, the firm’s Vantage module reviewed thousands of contracts to assess compliance with the EU’s Dora regulation for banks. Different AI tools serve different needs, reflecting the varied demands of legal practice.

Supporting Firm Integration

During the Allen & Overy-Shearman merger, Contract Matrix enabled rapid knowledge sharing across teams. Within two weeks, it was deployed to bridge systems and ease collaboration, illustrating AI’s role in managing complex organisational change.

Looking Ahead: AI Agents and Legal Expertise

The firm is co-developing AI agent modules that interact dynamically with lawyers. One agent assists antitrust teams by analysing M&A deal filings worldwide, requesting additional data as needed and refining its output. These agents use proprietary data and reflect expert input, ensuring high-quality, confidential results.

Experiments with AI “tournaments,” where agents trained in different legal areas negotiate to find solutions, are underway—pointing to new frontiers in AI-assisted legal analysis.

Regulation and Global Perspectives

Governments are increasingly viewing AI as an economic driver. However, regulatory approaches vary globally, from active regimes in some regions to more cautious “wait and see” stances elsewhere. The firm advises clients and governments on navigating these evolving landscapes.

For legal professionals interested in how AI can transform their work while maintaining compliance and ethics, exploring tailored AI training and certification can be valuable. Resources like Complete AI Training’s courses for legal professionals offer practical skills aligned with current industry needs.