Employees are already using AI tools, with or without permission. Clear guidance is better than hoping nothing goes wrong.

Where AI helps

– Drafting and editing emails, reports, and proposals

– Summarizing long documents and meetings

– Brainstorming and outlining

– Explaining code or spreadsheets

– Translating and rephrasing text

The main risks

– Data leakage: pasting confidential or personal information into a public tool

– Inaccuracy: AI can sound confident and still be wrong

– Copyright and originality: outputs may resemble existing content

– Bias and tone problems

– Unapproved tools: accounts created with personal emails and no oversight

– Privacy law: personal information about customers or employees brings legal obligations, including under Law 25

Practical rules to set

1. Approve specific tools, preferably business versions with clear data protections and no training on your data

2. Never enter passwords, client confidential data, personal information, or financial details into unapproved tools

3. Review every output. A person is responsible for what is sent or published.

4. Verify facts and figures against reliable sources

5. Disclose AI use where clients or regulations expect it

6. Keep records of how AI contributed to important work

Build the policy

A one-page AI use policy covers approved tools, prohibited data, review requirements, and who to ask. Share it and train staff.

Choose business-grade tools

Look at data residency, access controls, administrative oversight, and contractual commitments about how your data is used.

Start with pilots

Test with a small group, learn what works, and then expand.

DBR Soft can help you write an AI policy and deploy approved tools securely. Contact us for information.