AI Usage Policy

Customise the [bracketed fields] before adopting as your firm’s policy.

1. Purpose

This policy sets out how team members at [Firm Name] may use artificial intelligence (AI) tools — including large language models (LLMs) such as Claude and ChatGPT — in their day-to-day work. It ensures AI is used responsibly, professionally, and in a way that protects our clients, tenants, and our firm.

2. Scope

This policy applies to all team members, contractors, and representatives of [Firm Name] who use AI tools in connection with their role.

3. Approved Tools

The following tools are approved for use under this policy:

• [e.g. Claude by Anthropic — claude.ai]
• [e.g. ChatGPT by OpenAI — chatgpt.com]
• [Add or remove as appropriate]

Any other AI tool must be approved by [Manager/IT/Compliance] before use.

4. Acceptable Use

AI tools may be used for:

• Drafting tenant and owner communications for human review before sending
• Summarising documents, inspection reports, and meeting notes
• Brainstorming strategies, retention approaches, and marketing ideas
• Creating template notices, checklists, and procedure documents
• Explaining legislation, lease clauses, and industry concepts

5. Prohibited Uses

AI tools must not be used to:

• Send communications directly to tenants or owners without human review
• Make final decisions on legal, financial, or compliance matters
• Store or process sensitive personal data beyond what is necessary for the task
• Generate content that could mislead clients or misrepresent the firm
• Bypass mandatory approval or escalation processes

6. Data Handling

When using AI tools, team members must:

• Avoid pasting full tenant names, addresses, and financial details where not required for the task
• Use anonymised or placeholder data where possible (e.g. “Tenant A”, “Property X”)
• Not upload confidential documents to AI tools without management approval
• Be aware that information entered into AI tools may be used to improve those tools unless enterprise settings are configured to prevent this

7. Human Review Requirements

The following AI outputs must always be reviewed by a human before use:

• All communications sent to tenants, owners, or external parties
• Any document containing legal language or compliance obligations
• Financial summaries, reports, or analysis
• Content referencing specific legislation or regulatory requirements

8. Mandatory Escalation Triggers

AI must not be the sole basis for decisions in the following situations — escalate to [Manager] before proceeding:

• Tenant eviction or termination of tenancy
• Disputes involving legal proceedings or tribunal
• Any situation involving a vulnerable tenant
• Compliance breaches or suspected fraud

9. Team Responsibilities

10. Policy Review

This policy will be reviewed by [Manager] every [6/12] months, or sooner if a significant change in AI capability or regulation requires it. Last reviewed: [Date].