Your team already uses AI. Now prove it’s governed.
Strac Comply finds the AI apps your team connected to Google Workspace and Microsoft Entra ID, and records who approved each one, and why. With Security-First, Strac also keeps sensitive data out of ChatGPT, Claude and MCP tools.
14-day free trial. No credit card.
See it in 75 seconds
Watch how Strac finds the AI tools and agents your team uses, and protects the data going in.
AI showed up before your AI policy did.
Shadow AI
People sign in to AI note-takers, slide makers and chat tools with their work Google or Microsoft account. Some of these apps can read mail, calendars and Drive.
Sensitive data in prompts
Customer records, API keys and health data get pasted into ChatGPT and Claude every day. A policy doesn’t stop that.
Auditors and buyers are asking
Security questionnaires now ask which AI tools you use, who approved them, and how you keep customer data out. “We have a policy” is a weak answer.
From “what AI do we use?” to audit evidence.
- Step 1
Find
Connect Google Workspace or Microsoft Entra ID. Strac Comply lists the apps people granted access to their work accounts and flags the known AI tools, with the scopes each one holds and how many people granted it.
- Step 2
Decide
Approve each app, mark it not approved, or accept the risk, with your reason on record. Strac Comply keeps who decided and when. Remove a Google Workspace app’s access in the same step.
- Step 3
Protect and prove
Move an approved app into your vendor list, and its vendor review counts toward SOC 2 and ISO 27001. On Security-First, Strac blocks or redacts sensitive data before it reaches ChatGPT, Claude or an MCP tool.
What you see in Strac Comply and Strac.
AI apps connected to company accounts
Google Workspace · Microsoft Entra IDChatGPTAI
drive.readonly · 14 people
ClaudeAI
gmail.readonly, drive.readonly · 9 people
AI note-takerAI
calendar, drive · 6 people
AI slide makerAI
drive, contacts · 1 person
AI note-taker not approved · ada@example.com
6 Google Workspace grants revoked: “Records customer calls. Not on our approved vendor list.”
AI apps connected to company accounts, with risk level and approval status.
Recent AI detections
ChatGPT
prompt
Claude
prompt
Cursor
MCP tool call
Google Drive
shared with AI app
Recent AI detections in Strac: an API key blocked in ChatGPT, PHI redacted in Claude, an AWS secret blocked in an MCP tool call. Included in Security-First.
Governance you can show, not just write.
| What you need | Policy-only AI governance | Strac Comply |
|---|---|---|
| Which AI apps can reach company data | You ask around | Found from Google Workspace and Entra ID |
| Approvals | A spreadsheet | Recorded per app, with who, when and why |
| Removing access | Each person, by hand | One step for a Google Workspace app, with the reason kept |
| Sensitive data going into AI | Training and hope | Blocked or redacted in the browser and MCP tools by Strac (Security-First) |
| Evidence for SOC 2 and ISO 27001 | Screenshots | An approved app joins your vendor list, and its vendor review counts as evidence |
Strac has protected sensitive data for enterprises for five years. Strac Comply brings that same care to your compliance program.
Counts toward the controls you’re already working on.
Move an AI app you approve into your vendor list in one click. Its vendor review then counts as evidence for:
- SOC 2 CC9.2
- ISO 27001 A.5.19
- ISO 27001 A.5.23
- HIPAA 164.308(b)
You can also review Google Workspace app grants in the same access review as your accounts. The closed review is filed as your SOC 2 access review evidence (CC6.3).
Included in Security-First
AI Governance, an AI Pen Test and Strac DLP for Slack, Google Workspace and AI tools, plus everything in Certified: your SOC 2 Type II audit and a human vCISO. $12,995 a year for teams up to 10 people.
Straight answers
Does Strac read our prompts?
AI app discovery does not read prompts. It reads the app grants in Google Workspace and Microsoft Entra ID.
Prompt inspection happens only when you turn on Strac DLP, included in Security-First, and then only to block or redact sensitive data.
Which AI tools do you cover?
Every app connected through Google Workspace or Microsoft Entra ID, with the known AI tools flagged at the top. With Strac DLP on Security-First, also ChatGPT and Claude in the browser and MCP connections on developer laptops.
Is this the same as an AI policy?
No. A policy says what people should do. Strac Comply shows which AI apps can reach company data and keeps a decision on each one, with the reason and who made it.
That record is what an auditor or a buyer asks for when they ask how your policy is followed.
Do I need the full Strac platform?
No. Security-First covers Slack, Google Workspace and AI tools. Endpoints and cloud data are available separately.
Get AI governance and SOC 2 from one platform.
14-day free trial. No credit card. Plans from $4,995 per year. See pricing