Responsible AI AI your compliance team can supervise

We’re adding AI features to help your teams identify items that need review. Each feature follows the controls below.

  • Off by default
  • Processed in your country
  • Your data is never used to train AI models
  • Human in the loop

Our AI principles

These principles apply to every AI feature across InvestorCOM products, including new releases.

Your firm controls AI use

AI features are off by default. Your firm can enable or disable them at any time.

Role-based permissions control each user’s access.

Your data isn’t used to train AI

We send the AI only what a feature needs, with identifying details removed. It’s processed in your firm’s country, Canada or the US. It’s never used to train AI models.

Human in the loop

AI flags items and suggests actions. A person at your firm makes the decision. We retain an encrypted decision record in our SOC 2-audited environment.

Supervision AI Agent: data use and controls

The details below apply to the Supervision AI Agent.

Supervision AI Agent

Data and controls
What it doesFlags recommendations for review and puts the at-risk items at the top of your queue.
Who decidesYour supervisor.
Data sent to AIIdentifying details are removed before processing.
Processed byMicrosoft Azure, in your firm’s country. Not shared with the company that built the model.
Used for trainingNo.
Kept by the AI providerOnly content the AI provider’s automated systems flag for abuse monitoring, in the same region, for up to 30 days.
Kept by our review engineAn encrypted working copy, deleted within 7 days.
Your audit recordKept, encrypted, for your regulatory retention period.
Off by defaultYes. Your firm turns it on.

Review process

  1. 1. Recommendation submitted
  2. 2. Identifying details removed
  3. 3. AI flags items for review
  4. 4. Supervisor review and decision
  5. 5. Decision recorded

Audits and standards

Our current audit coverage, planned certification and standards under evaluation.

SOC 2 Type 2

In place

Covers the environment where your AI records are kept.

ISO/IEC 42001

Certification planned

The international standard for managing AI.

AIUC-1

Under evaluation

An independent standard for the security and safety of AI agents.

For our platform-wide controls, see Security.

Have questions about how we use AI, or want to see it in action?