AI Use Policy
Effective: 30 March 2026
Compliant with BSB Core Duties and Bar Council Guidance on Generative AI (November 2025)
1. Purpose and Scope
This policy governs my use of artificial intelligence (AI) tools, including large language models (LLMs) such as ChatGPT, Google Gemini, Perplexity, Harvey, and Microsoft Copilot—in my practice as a sole-trading barrister. It ensures compliance with:
BSB Core Duties (particularly CD1: duty to the court; CD2: best interests of client; CD3: honesty/integrity; CD6: confidentiality; CD7: competent work)
Bar Council's Updated Guidance on Generative AI (November 2025)
GDPR/Data Protection Act 2018
Legal Professional Privilege (LPP) obligations
AI use is permitted only as an augmenting tool; ultimate responsibility for all legal work remains with me.
2. Permitted Uses of AI
AI may be used responsibly for:
2.1 Approved Tasks
Drafting initial outlines or first drafts of non-substantive documents (e.g., skeleton argument structure, correspondence templates)
Summarising well-defined areas of law (where I have expertise to evaluate accuracy)
Brainstorming arguments or identifying potential issues
Proofreading and editing for clarity/grammar
Administrative tasks (calendar management, time tracking, document organisation)
2.2 Conditions for Use
I must understand the capabilities and limitations of each AI tool before
AI outputs must be verified, reviewed, interpreted, and contextualised for accuracy
AI must never replace human judgment or professional expertise
3. Prohibited Uses of AI
3.1 Absolute Prohibitions
Submitting AI-generated content to the court without independent verification
Citing AI-generated case law, statutes, or legal authorities without confirming:
The case exists
The citation is accurate
The legal proposition is correctly stated
The case remains good law
Inputting confidential client information, privileged material, or personal data into public AI tools without adequate safeguards
Using AI to draft core legal arguments without substantive independent analysis
Relying solely on AI for legal research without checking authoritative sources (Westlaw, LexisNexis, Inns of Court libraries)
4. Verification Obligations
For all AI-assisted work, I will:
Legal citations: Confirm case exists, citation accurate, proposition correct, case still good law
Legal analysis: Cross-check against authoritative databases (Westlaw, LexisNexis, BAILII)
Factual assertions: Verify against primary sources (evidence, documents, statutory text)
AI confidence: Never rely on AI's apparent confidence - verify independently regardless
5. Data Protection & Confidentiality
5.1 GDPR Compliance
Ensure lawful basis for processing any personal data input to AI systems
Verify AI providers offer adequate data protection commitments
Review AI tool terms to confirm they permit processing of client matter data
Document decisions about AI use where personal data is involved
5.2 Confidentiality Measures (Core Duty 6)
Never input client names, case details, privileged communications, or sensitive personal data into public LLMs
Use enterprise/private AI instances with data protection agreements where client data must be processed
Anonymise data before input where possible (remove identifying information)
Maintain legal professional privilege at all times
6. AI Tool Assessment & Registration
6.1 Approved Tools Register
I maintain a register of assessed AI tools.
6.2 Assessment Criteria Before Using New Tools
Provider's data protection commitments (GDPR compliance)
Data storage location (UK/EEA preferred)
Whether outputs are used for model training
Security vulnerabilities
Reputation and reliability in legal context
7. Competence & Training (Core Duty 7)
7.1 Knowledge Requirements
I will maintain understanding of:
AI tool capabilities and limitations
Verification best practices
Data protection implications
Ethical obligations in AI-assisted practice
7.2 CPD Commitment
Complete annual training on AI tools and ethical use
Stay current with regulatory developments (Bar Council/BSB guidance updates)
Understand that competent use of AI is now a professional competence expectation
8. Disclosure to Clients and Court
8.1 Client Disclosure
Inform clients when AI will be used in their matter (where material to engagement)
Explain limitations of AI and that human oversight is maintained
Address any client concerns about AI use
8.2 Court Disclosure
Current Bar Council guidance notes no mandatory disclosure requirement yet but similar requirements may develop
If directly asked by the court whether AI was used, I will answer truthfully (Core Duty 3: honesty and integrity)
Follow any court-specific standing orders on AI disclosure
9. Record-Keeping & Audit Trail
I will maintain records of:
Which AI tools were used for each matter
What AI-generated content was produced
Verification steps undertaken (critical for demonstrating competence)
Client consent where AI use is material
Data protection assessments for AI tools used
Records retained for minimum 6 years (professional indemnity requirements).
10. Breach Response Protocol
If I suspect AI has produced inaccurate, hallucinated, or fabricated content:
Immediately halt use of that output
Conduct full verification using authoritative sources
If AI-generated content was submitted to court/client:
Notify instructing solicitor immediately
File correction/corrigendum to court if necessary
Document the error and remediation steps
Report serious breaches to BSB if required (Core Duty 9: open and cooperative with regulators)
Review and update verification procedures to prevent recurrence
11. Policy Review
This policy will be:
Reviewed annually (minimum)
Updated immediately following BSB/Bar Council guidance changes
Revised after any AI-related error or regulatory development
Next review date: 30 March 2027