By João Rodrigues · Better Skills · Published 23 July 2026
This is the English overview page of AI Act Portugal, a Portuguese-language platform about the EU AI Act. The full guides and free tools are available in Portuguese.
The EU AI Act in Portugal — English Guide
What Regulation (EU) 2024/1689 requires, when it applies after the 2026 Digital Omnibus, and how Portugal supervises it
What is the EU AI Act?
The Artificial Intelligence Act — Regulation (EU) 2024/1689 — is the world's first comprehensive legal framework for artificial intelligence. It entered into force on 1 August 2024 and applies in phases across all EU Member States, including Portugal, without the need for national transposition. The Act follows a risk-based approach: it bans a closed list of unacceptable practices, imposes strict requirements on high-risk systems, sets transparency duties for AI that interacts with people or generates content, and regulates general-purpose AI (GPAI) models.
Updated timeline after the Digital Omnibus (2026)
In June 2026 the EU gave final approval to the Digital Omnibus on AI (European Parliament vote on 16 June; Council approval on 29 June), a package that postponed the high-risk obligations and simplified several rules. The current timeline:
| Date | What applies |
|---|---|
| 2 Feb 2025 | Prohibited practices (Art. 5) and AI literacy (Art. 4) — in force |
| 2 Aug 2025 | GPAI model obligations, governance and penalties — in force |
| 2 Aug 2026 | Transparency obligations (Art. 50): chatbots, deepfakes, AI-generated content. Content marking (Art. 50(2)) has a transition until 2 Dec 2026 for systems already on the market |
| 2 Dec 2026 | New Art. 5 prohibition (non-consensual intimate imagery / CSAM generators) and end of the watermarking transition |
| 2 Aug 2027 | Regulatory sandboxes operational in every Member State |
| 2 Dec 2027 | High-risk systems under Annex III (postponed from 2 Aug 2026) |
| 2 Aug 2028 | High-risk AI embedded in regulated products under Annex I (postponed from 2 Aug 2027) |
How Portugal supervises the AI Act
Portugal designated ANACOM (the national communications authority) as the coordinating supervisory authority on 19 September 2025, working alongside 14 sectoral authorities — including the Bank of Portugal's counterparts for finance (IGF), health (ERS), labour (ACT), energy (ERSE), media (ERC), law enforcement (PJ, IGAI) and education (IGEC). The national sanctions regime implementing the EU penalty framework (fines up to €35 million or 7% of global turnover for prohibited practices; €15 million or 3% for other infringements) is still being finalised, but the EU obligations apply directly in the meantime.
Portugal also approved a national AI agenda for 2026-2030 (ANIA), which includes the Portuguese large language model "Amália", a national AI Factory and the regulatory sandbox required by the Act.
What should companies operating in Portugal do now?
- Check the prohibitions and literacy duties — both have applied since February 2025 to providers and deployers of any AI system;
- Prepare for transparency by 2 August 2026 — customer-facing chatbots must disclose they are AI; deepfakes and synthetic content must be labelled;
- Plan high-risk compliance for December 2027 — risk management, data governance, technical documentation, human oversight and conformity assessment for Annex III systems;
- Document everything — supervisory authorities will expect demonstrable, written compliance.
Disclaimer: this page is informational and does not constitute legal advice. For decisions with legal impact, consult qualified counsel and the official text of the Regulation on EUR-Lex.
Free tools and resources (in Portuguese)
AI Act Portugal offers free interactive tools — an AI risk classifier, deadline calendars and trackers, a transparency checker, a GPAI/FLOPs checker, a 78-question compliance audit and more than 100 document templates. They are available in Portuguese: