EU AI Act Compliant Productivity Intelligence Platform — built for August 2026

gStride is designed with EU AI Act categories in mind from the architecture up. AI inferences are framed as recommendations to humans rather than autonomous decisions, surveillance components (screenshots, keystroke capture) are configurable and off by default, no emotion or stress inference exists in the product, and the human-oversight workflow is documented. Risk-management documentation, technical documentation of inferences, and post-market monitoring work are in progress ahead of the August 2026 enforcement date. We do not claim to be certified or approved by the EU; AI Act readiness is an architecture and operating discipline, not a feature. The full position is in the AI Act compliance checklist. Verify your specific obligations with counsel for your jurisdiction.

The EU AI Act became law in 2024 and is being phased in. The high-risk-system obligations relevant to workplace AI begin to apply from August 2, 2026. By that date, providers and deployers of high-risk employment AI must have transparency notices, human oversight, conformity assessment, risk management, logging, and post-market monitoring in place. The exact dates and staged enforcement timeline are subject to revision in implementing regulations; verify with counsel for your jurisdiction.

The category most affected is the classic time tracker with bolted-on AI features. To the extent vendors ship AI-driven productivity scoring, AI activity classification used in HR review, or AI ranking used for promotion or termination, those features are likely in Annex III scope. Keystroke and mouse-jitter wellbeing or engagement inferences sit higher up in Article 5 prohibited territory. Webcam emotion or video sentiment inferences in employment are likewise prohibited. EU customers should request a written AI Act readiness statement and an Annex III scope letter from each incumbent vendor ahead of August 2026. See gStride vs Hubstaff and gStride vs Time Doctor for two vendor-specific reads.

Where gStride uses AI (focus signal, billing variance, idle classification, capacity drift, meeting overhead patterns), the output is presented to a human reviewer as a recommendation with full override authority and the option to dismiss without consequence. The AI is never the final decision-maker on a performance review, promotion, termination, or material task allocation. This design choice maps onto Article 14 human-oversight requirements and matters because productivity scoring that auto-feeds into a performance-improvement plan without a meaningful human review step is non-compliant by design.

No. The category is excluded by design, not by toggle. gStride does not infer mood, stress, engagement, or wellbeing from keystroke cadence, mouse movement, webcam, or microphone in any employment context. Emotion-recognition AI in the workplace is restricted under Article 5 of the AI Act outside narrow safety and medical exceptions. The wellbeing or burnout reframing some vendors use does not change the legal classification, so we exclude the category entirely. See the AI idle detection vs keystroke logging writeup for the architectural distinction.

Customers get a deployer kit that includes a template worker-notification notice for Article 13 transparency, a human-oversight workflow default with role-mapping, a logging policy with retention windows, and onboarding guidance on per-feature opt-in monitoring. The kit is sized for EU AI Act employee transparency and works alongside the customer's own DPIA. The platform is configurable to run productivity intelligence without any surveillance capture at all. See also the GDPR-compliant employee monitoring checklist for the data-protection layer that sits underneath. Verify final deployer obligations with your counsel for your jurisdiction and works-council situation.

In jurisdictions with works councils (Germany, Austria, Netherlands, France in scope industries, and others), AI-based monitoring is a consultation-required deployment under existing labour law as well as the AI Act transparency layer. gStride's deployer kit is designed to support the works-council briefing rather than work around it — the configurability of each monitoring feature, the recommendation-not-decision framing of inferences, and the absence of Article 5 features are the same talking points the council needs to hear. Skipping the council is the largest source of legal exposure in this category and is not a path we will help customers take.