AI Policy

How Arctex's AI agent features actually use your data. Last updated 2026-10-08.

What AI agents can access

When you use an AI agent feature, it reads and acts on your own business's data — contacts, calendar, pipeline, reviews, and similar records — to do the specific task it's running. This access is enforced the same way every other part of the platform enforces it: an agent running for your business cannot read or write another business's data.

Where that data goes

The relevant data is sent to our AI model provider (currently Anthropic) to generate the agent's response. It's retained as part of the record of that agent run, the same way other activity on the platform is recorded, so you can review what an agent did and why.

Human review

AI-generated output is labeled as AI-generated wherever it's shown for review, and where an agent drafts something that will be published externally (for example, a social media post), that's flagged before it goes out. You're responsible for reviewing AI-drafted content before relying on it or publishing it — see our Acceptable Use Policy.

What we ask you not to do

Don't input another person's sensitive personal, health, or financial information into an AI feature unless you have a lawful basis to process it. Full restrictions are in our Acceptable Use Policy.

Changes

As we add real data-retention controls or change AI providers, we'll update this page to reflect it.