# SRA AI Guidance & UK Legal AI Regulation

**Date:** 2026-06-27  
**Research cycle:** 1 — Legal Intelligence  
**Scope:** SRA guidance on AI use by solicitors, UK court guidance, Divisional Court / King’s Bench warnings, and Law Society positions.

## 1. Solicitors Regulation Authority (SRA)

### AI Risk Outlook (November 2023)

The SRA published a **Risk Outlook report: “The use of artificial intelligence in the legal market”** setting out its view of the benefits and risks of AI in legal services, including: [^1]

- Potential benefits: efficiency, cost reduction, improved access to legal services.
- Key risks: **accuracy, confidentiality, transparency, bias, accountability** and **client understanding**.
- The SRA uses the **ICO definition** of AI as algorithmic systems that solve complex tasks previously requiring human intelligence.

### Regulatory Approach

- The SRA does not ban AI use by solicitors. Instead, it expects firms to comply with the existing **SRA Standards and Regulations**, in particular:
  - **Principle 1:** uphold the rule of law and proper administration of justice.
  - **Principle 2:** act with integrity.
  - **Principle 3:** not allow independence to be compromised.
  - **Principle 4:** act in the best interests of each client.
  - **Principle 5:** provide a proper standard of service.
  - **Principle 6:** behave in a way that maintains public trust.
  - **Principle 7:** comply with legal and regulatory obligations.
- Core duties affected by AI use include: **client confidentiality, competence, supervision, disclosure of costs and confidentiality risks**, and **conflict checking**.

### First AI-Authorised Law Firm: Garfield.Law

In **May 2025**, the SRA authorised **Garfield.Law Ltd** as the **first purely AI-driven law firm** to provide regulated legal services in England and Wales.[^2]

- The firm provides an AI-powered litigation assistant focused on small-claims debt recovery for SMEs.
- SRA approval came with strict safeguards:
  - client confidentiality protection;
  - conflict-avoidance mechanisms;
  - user approval at each stage;
  - controls preventing AI from proposing case law (addressing hallucination risk).
- The authorisation is treated as a **landmark moment** for legal services regulation in the UK.

### Compliance Tips for Solicitors (2026)

The SRA maintains a “Compliance tips for solicitors regarding the use of AI and technology” resource, which cross-references UK Government public-sector AI guidance and highlights the need to safeguard confidentiality and maintain trust, privacy and compliance.[^3]

---

## 2. Courts and Judiciary

### Judicial AI Guidance (October 2025)

In **October 2025**, the Judiciary issued revised **“Artificial Intelligence (AI) – Judicial Guidance”**. [^4]
- Lord Justice Birss, Lead Judge for AI, stated that AI use by the judiciary must be consistent with its overarching responsibility to protect the integrity of the justice system.
- The guidance is deliberately conservative: AI can **assist**, not act as a **shortcut**.
- It addresses risks of generative AI, including bias, hallucination, and confidentiality.

### King’s Bench / Divisional Court Warnings

In **June 2025**, in **R (London Borough of Haringey) [2025] EWHC 1383 (Admin)**, the President of the King’s Bench Division issued a wake-up call to the profession about the risks of AI in court documents.[^5]
- Lawyers were warned they could face **criminal charges** if they rely on **fictitious AI-generated case law**.
- The Bar Council summarised the ruling as a clear signal that practitioners remain personally accountable for material filed with the court. [^6]

### Confidentiality / Privilege Warning (2026)

A Norton Rose Fulbright note from April 2026 warns that use of open-source GenAI may **waive confidentiality and legal professional privilege** if sensitive material is input into public models.[^7]

---

## 3. Law Society of England and Wales

### Generative AI: The Essentials

The Law Society publishes **“Generative AI: the essentials”** guidance, acknowledging that GenAI has created "boundless opportunities" but also brings **new technology and data risks**.[^8]

Key themes:
- Understand the technology’s limitations.
- Maintain **client confidentiality** and **data protection** obligations.
- Ensure proper **supervision** and **competence**.
- Keep clients informed where AI is used.
- Review outputs critically to avoid errors and hallucinations.

### Policy Position

The Law Society argues the UK Government should adopt a **balanced approach to AI regulation**: support innovation while safeguarding ethical standards and the rule of law.[^9]

---

## 4. Other Relevant Regulators & Frameworks

| Body | Relevance |
|------|-----------|
| **Information Commissioner’s Office (ICO)** | AI definition used by SRA; data-protection obligations when AI processes personal/confidential data. |
| **UK Government AI public-sector guidance** | Cross-referenced by SRA compliance resources. |
| **Council of Europe Framework Convention on AI** | First legally binding international AI treaty; UK context for human rights, democracy and rule of law. |

## 5. What This Means for Sovael Legal

1. **Compliance-first positioning is mandatory.** Any AI service must map directly to SRA Principles.
2. **Confidentiality is non-negotiable.** Client data should not be entered into public/open-source models without explicit safeguards.
3. **Human accountability remains absolute.** Outputs must be verified; lawyers remain responsible for court submissions.
4. **Transparency with clients** on AI use is both an ethical and commercial advantage.
5. **Garfield.Law proves the model is licensable** — but only with rigorous controls, making AI governance advisory a natural Sovael service.

## Sources

[^1]: SRA, “Risk Outlook report: The use of artificial intelligence in the legal market,” 20 November 2023 — https://www.sra.org.uk/sra/research-publications/artificial-intelligence-legal-market/
[^2]: SRA, “SRA approves first AI-driven law firm,” May 2025 — https://www.sra.org.uk/sra/news/press/garfield-ai-authorised/; LexisNexis, “SRA authorises first AI-driven law firm,” 2025 — https://www.lexisnexis.co.uk/legal/news/sra-authorises-first-ai-driven-law-firm-to-provide-regulated-legal-services
[^3]: SRA, “Compliance tips for solicitors regarding the use of AI and technology,” February 2026 — https://www.sra.org.uk/solicitors/resources/innovate/compliance-tips-for-solicitors/
[^4]: Judiciary UK, “Artificial Intelligence (AI) – Judicial Guidance (October 2025),” 31 October 2025 — https://www.judiciary.uk/guidance-and-resources/artificial-intelligence-ai-judicial-guidance-october-2025/
[^5]: IBA, “Technology: UK judge warns lawyers about risks of AI use in court,” 28 July 2025 — https://www.ibanet.org/Technology-UK-judge-warns-lawyers-about-risks-of-AI-use-in-court
[^6]: Bar Council, “AI in law: breaking down the latest developments,” 4 July 2025 — https://www.barcouncil.org.uk/resource/ai-in-law-breaking-down-the-latest-developments.html
[^7]: Norton Rose Fulbright, “Court guidance that use of open-source AI waives confidentiality and legal professional privilege,” April 2026 — https://www.nortonrosefulbright.com/en/inside-disputes/blog/202604-court-guidance-that-use-of-open-source-ai-waives-confidentiality-and-legal-professional
[^8]: The Law Society, “Generative AI – the essentials” — https://www.lawsociety.org.uk/topics/ai-and-lawtech/generative-ai-the-essentials
[^9]: The Law Society, “AI and lawtech: government policy and regulation,” 7 May 2026 — https://www.lawsociety.org.uk/Topics/AI-and-lawtech/Whats-changing/AI-lawtech-policy