Step 9 of 16
What data to never put into AI — company policies, personal data, financial info. Practical guidelines for Thai workplace context.
ข้อมูลอะไรไม่ควรใส่เข้า AI เด็ดขาด — ข้อมูลบริษัท ข้อมูลส่วนตัว ข้อมูลการเงิน
If you wouldn't post it on a public bulletin board, don't put it in a free AI tool.
Free AI tools may use your conversations to train models. Paid/enterprise tools usually have agreements that protect your data. Always check before you type.
| Data Type | Free AI | Paid/Enterprise | Safe Approach |
|---|---|---|---|
| Company strategy | Don't | Check policy | Anonymize first |
| Source code | Don't (public repos) | Check policy | Use private repos only |
| Meeting notes with names | Don't | Usually OK | Remove names |
| Financial reports | Don't | Check policy | Use percentages, not actuals |
| Client requirements | Don't | Check policy | Genericize details |
Before pasting work content into AI:
Original: "Our client Siam Commercial Bank wants to migrate
their core banking system from the legacy IBM mainframe
by Q3 2026 with a budget of 50M THB"
Anonymized: "A large Thai bank wants to migrate their core
system from a legacy mainframe by Q3 2026 with an 8-figure
budget. What are the key risks and recommended approach?"
You get the same quality of advice without exposing sensitive data.
If your team doesn't have an AI policy, start with these rules:
AI is useful but your data is irreplaceable. When in doubt, anonymize. Use paid tiers for sensitive work. And always remember: AI output needs human review before it reaches clients or stakeholders.