See · Prompt & response capture

Prompt
& response capture

PromptGlass records AI prompts, responses and file uploads in a searchable, encrypted store, with the level of detail set per policy: full content, redacted content, metadata only, or a hash. Investigators get the evidence they need; everyone else sees only what they are entitled to.

See it in a demo Read the docs

Outcomes

✓Defensible evidence for investigations
✓A record that supports AI audits
✓Privacy controls that satisfy works councils

What you get

Inside PromptGlass prompt & response capture

Four storage modes

Full, redacted, metadata-only or hashed, chosen per app, group or data class.

Customer-managed keys

Content is encrypted with your KMS or HSM keys in self-hosted deployments. Revoke the key and the content is unreadable.

Search and replay

Find every prompt that mentioned a customer, project code name or file hash, across every AI tool.

Legal hold and retention

Retention per data class, legal hold for investigations, and deletion that is provable.

Rollout

Four steps, no big bang

Start in observe-only mode and tighten controls as you learn what your organization needs.

1
Choose storage modes
Start with metadata-only and turn on content capture for selected apps or data classes.
2
Set retention
Keep metadata longer than content, and keep sensitive classes for the shortest time.
3
Assign access
Role-based access with four-eyes approval to reveal identities or content.
4
Search
Query by user, app, data class, topic or free text.