Kvíz spojený s průvodcem · Střední úroveň

How to Protect Confidential Data When Using AI at Work Quiz

Check classification, approved accounts, terms and retention before sharing confidential work material with an AI service.

Otázka 1 z 8

What should an employee check before entering confidential work data into an AI tool?

Náhled každé otázky8 otázky

Odpovědi a vysvětlení se zobrazují, jak budete v kvízu.

  1. What should an employee check before entering confidential work data into an AI tool?
    • Whether the answer will sound professional
    • The organization’s classification rules and approval for that exact tool/account
    • Whether the prompt is short
    • Whether the model has a friendly tone
  2. What does OpenAI state about business/API data and training by default?
    • It is retained for exactly 30 days
    • No service personnel can ever access it
    • It is used for training unless an admin deletes it
    • Business/API inputs and outputs are not used for training by default; retention still needs separate review
  3. Why does “not used for training” not resolve every data-protection question?
    • It means the data is immediately deleted
    • It does not establish retention, access, configuration, contract scope or policy approval
    • It proves every data type is approved
    • It ensures connected tools cannot receive content
  4. When preparing data for an approved AI tool, which redaction practice reduces exposure?
    • Remove unnecessary identifiers and check whether remaining details still expose sensitive facts
    • Replace vowels and assume anonymity
    • Keep all fields for more context
    • Use a personal account to avoid workplace controls
  5. What does NIST SP 1800-39 address?
    • A guarantee redacted data cannot be reidentified
    • A universal requirement to choose a vendor
    • Discovering, identifying and labeling sensitive unstructured data as a practice guide
    • Certification of all AI products
  6. If a service excludes customer data from model training, what can still vary?
    • Nothing; all data is immediately erased
    • Model accuracy only
    • Retention, access, integrations, contract scope and account configuration
    • Whether employees have policy obligations
  7. What should someone do if a tool is not approved for the data type?
    • Submit a small excerpt
    • Switch to a personal account
    • Assume public information means approval
    • Ask the designated security/privacy contact before sharing
  8. After confidential information is accidentally entered into an unapproved tool, what response is safest?
    • Delete the visible chat and assume all copies are gone
    • Follow the organization’s incident-reporting process promptly
    • Wait for a warning
    • Send the data to another provider