Employee GenAI usage policy for India that teams can actually follow
Last updated:
A ready to use policy outline you can adapt for your company. It focuses on two areas where teams make costly mistakes: sharing confidential data in ChatGPT style tools, and unclear ownership of AI generated work.
Offrd is currently available in English only for Indian businesses.
1. Scope and approved tools
Define who the policy covers and which GenAI tools are approved for work. Include employees, interns, contractors, and vendors working on company systems.
- Approved accounts and workspaces only, if your company provides them
- No connecting chat tools to company drives, email, ticketing, or code repos without written approval
- External facing outputs need review before they are shared
Helpful reference: Essential employment policies guide
2. Confidentiality rules for ChatGPT style tools
This is the part most teams need in plain language. Assume prompts and uploads may be stored by the provider.
Never paste or upload
- Client data, contracts, SOWs, pricing, invoices, bank details
- Source code, architecture diagrams, private repo content
- Passwords, API keys, tokens, certificates, or access links
- Employee personal data such as Aadhaar, PAN, bank, salary, medical records
Allowed with care
- Public information and generic templates
- Redacted samples where identifiers are removed and cannot be reconstructed
- Fully synthetic examples and fabricated test data
3. Copyright and ownership of AI generated work
Keep this consistent with your offer letter and employment terms. The simplest rule is that work product created in the course of employment is owned by the company, even when AI helps draft it.
- Employees must not claim personal ownership of company deliverables due to AI assistance
- Do not accept tool terms on behalf of the company without approval
- Maintain proof of rights for any third party assets used in marketing
4. Code, licenses, and review controls
Treat AI output like code copied from the internet. It can be wrong, insecure, or licensed in ways you do not want.
- Mandatory code review before merge or release
- Run security scanning and dependency checks where your process supports it
- Avoid copying large blocks of AI output into production without validation
- Document key design decisions for audit and maintenance
For HR teams and founders, these checks mirror the same idea as keeping clean employee records for audits.
Labour Code 2025 readiness and everyday HR records
Many Indian companies are updating appointment letters, salary structure, and digital records as Labour Code 2025 requirements and interpretations evolve. Your GenAI policy should support that effort, not create new risk.
- Do not paste payslips, wage registers, PF or ESI details into public chat tools
- Keep policy drafts, appointment letters, and salary breakup formats inside your company systems
- For any statutory wording, rely on your CA or legal advisor for final review
Related pages: New Labour Code 2025 overview and Free payslip generator
5. Training, enforcement, and reporting
Policies work when people know what to do when they make a mistake. Add a simple reporting path and keep the steps short.
- Report accidental disclosure immediately to your IT or security contact
- Save final work in company systems, not in chat history
- Repeat issues trigger access restrictions and HR action
If you want a one page checklist for managers, you can add it as an appendix.
6. Template add ons for HR
If your HR team is building a full policy set, these pages are useful companions.
FAQ
Quick answers HR teams in India usually need when rolling this out.
Can employees use ChatGPT for work tasks
Who owns AI generated work created during employment
Should we ban AI tools completely
Does Offrd support HR letters and document templates
Disclaimer: This page is informational and does not replace professional legal advice. For statutory interpretation and company specific decisions, rely on your CA or legal advisor.