Does Jarvis Comply With the EU AI Act?

Yes, Jarvis (getjarvis.eu) is built to be EU AI Act aligned. As a general-purpose AI assistant it follows the Act's emphasis on transparency, data governance, and user control: you can see what Jarvis remembers and clear it, screen capture is on-demand and visible, actions are traceable in connected apps, and the system never trains on your data. Hosting is in the EU, dovetailing with GDPR. Jarvis routes tasks to frontier models from Anthropic, OpenAI, and Google, frontier models whose own providers carry obligations under the Act. Jarvis is the screen-aware desktop assistant (floating bar, Cmd+/ on macOS, Ctrl+/ on Windows) for macOS 12+, Windows 10+, and Linux. Founded by Mihai Matei in Romania, launched April 30, 2026. Free to start, then $16/month.

The EU AI Act introduces a risk-based framework with strong themes of transparency, accountability, and data governance. For an assistant like Jarvis the relevant duties center on being clear about what the system does, letting users understand and control automated behaviour, and handling data responsibly. Jarvis addresses these through inspectable memory, deliberate rather than covert screen capture, traceable connector actions, and EU data residency that complements its GDPR posture.

A guiding AI Act principle is that people should understand and stay in control of AI acting on their behalf. Jarvis keeps the human in the loop: you trigger screen awareness explicitly with the hotkey, you approve each connector via OAuth, and actions like sending email or posting to Slack are visible in those apps. The memory store is open to inspection and editing, so you can see and correct what the system believes about you, which is exactly the kind of oversight the Act encourages.

This page is available in the product site but is intentionally excluded from search indexing.

Privacy & data