The AI Act is the European regulation on artificial intelligence, approved in 2024 and with phased entry into force through 2026-2027. It classifies AI systems into four risk levels (unacceptable, high, limited, minimal) and imposes proportionate obligations. High-risk AI systems — which include many use cases in HR, health, credit access and critical infrastructure — require technical documentation, risk management, training-data quality, human oversight, robustness, transparency, and auditable records.
In a Portuguese corporate environment, the AI Act especially affects: predictive models in HR (flight risk, candidate evaluation), recommendation systems that affect decisions about people, and automation of decisions previously made by humans. Non-compliance can result in fines of up to 7% of annual turnover, heavy even by European standards.
INFOS develops and implements AI following the principles of the AI Act from day one: explainable models (no opaque black boxes), complete technical documentation, auditable training data, human oversight at decision points, usage logs. The AI layer of pplPortal and Document Management with cognitive capture are examples of INFOS systems designed for compliance from the outset.