Artificial intelligence news.
Straight from the original announcements.

Fiddler AI Sets Out EU AI Act Compliance Roadmap for Organizations

November 17, 2026By the AI News Bites Editorial TeamSource: Fiddler AI
Fiddler AI Sets Out EU AI Act Compliance Roadmap for Organizations
Image: Fiddler AI

Fiddler AI has published an eight-step roadmap for organizations preparing to govern artificial intelligence under the EU AI Act. The guide says companies should begin with a complete inventory of their AI systems, including workplace tools such as ChatGPT and Microsoft Copilot, before assigning risk tiers or compliance responsibilities.

Key points
  • Fiddler AI recommends an eight-step roadmap for EU AI Act governance.
  • Annex III high-risk obligations now apply from 2 December 2027.
  • Annex I product obligations now apply from 2 August 2028.
  • AI inventories should include employee use of ChatGPT and Microsoft Copilot.
  • Organizations should document whether they are providers or deployers.

High-risk deadlines have moved

Fiddler AI’s roadmap reflects changes introduced by Regulation (EU) 2026/1744, known as the Digital Omnibus on AI. The regulation entered into force on 27 July 2026 and postponed some high-risk obligations, while prohibitions, AI literacy requirements and obligations for providers of general-purpose AI kept their original dates.

Prohibitions under Article 5 and the AI literacy duty have applied since 2 February 2025. National authorities can enforce the literacy duty from 2 August 2026. The guide says obligations for high-risk systems covered by Annex III now apply from 2 December 2027, while product-embedded systems covered by Annex I move to 2 August 2028. High-risk systems intended for use by public authorities have an 2 August 2030 date under Article 111(2).

Fiddler AI warns that materials written before July 2026 may still list August 2026 as the high-risk deadline. It recommends checking dates against the consolidated text of 27 July 2026 before using them in board reports or project plans.

Inventory and role decisions come first

The first step is to record every AI system an organization builds, buys, embeds in a product or allows employees to use. The inventory should include coding agents and workplace use of ChatGPT or Microsoft Copilot, as well as each system’s owner, intended purpose, lifecycle status, supplier, role, risk tier, EU nexus and personal data processing.

Fiddler AI also highlights the problem of shadow AI. It cites a Gartner survey of 302 cybersecurity leaders conducted from March to May 2025, in which 69% of organizations suspected or had evidence of employees using prohibited public generative AI.

Organizations should then determine whether they are a provider, deployer or downstream provider for each system. A company can become a provider of a high-risk system if it places its name or trademark on it, makes a substantial modification that keeps it high-risk, or changes its intended purpose so that it becomes high-risk. Fiddler AI recommends recording the reasoning behind each role decision in the inventory.

Risk screening and governance responsibilities

Fiddler AI says prohibited practices should be screened first because those rules have applied since February 2025 and carry the highest fines. Examples listed in the guide include harmful manipulation, social scoring, untargeted facial-image scraping and workplace emotion recognition.

High-risk systems include safety components of regulated products and stand-alone systems in areas such as biometrics, critical infrastructure, education, employment, essential services, law enforcement, migration, and justice and democratic processes. The guide specifically identifies systems that “analyse and filter job applications” and credit-scoring systems as examples of uses named in Annex III.

The roadmap then calls for a cross-functional governance body involving legal, compliance, data protection, security, engineering and business owners. It should have authority to approve, block and retire AI systems. Fiddler AI also recommends an Article 4 AI literacy program and says completion records will provide evidence for national authorities. The company cites ISACA’s 2026 AI Pulse Poll, which found that 38% of organizations “have a formal, comprehensive AI policy,” compared with 28% in 2025.

For organizations, the guide’s message is that EU AI Act preparation starts with visibility and accountability. Existing controls from ISO/IEC 42001, GDPR, the NIST AI RMF and DORA can be reused, but providers and deployers still need to keep their roles, risk decisions and separate conformity assessment, CE marking and EU database registration steps clear.

Related news

Safety and ethics
Safety and ethicsNovember 12, 2026

OpenAI commits $2 million to research on AI safety for children

OpenAI has signed the Charter of Engagement for Safer AI for Children and committed $2 million to a scientific research network focused on children’s rights and…

UN launches toolbox to keep people accountable for AI in smart cities
BusinessNovember 6, 2026, 14:43

UN launches toolbox to keep people accountable for AI in smart cities

The United Nations has launched its first system-wide toolbox for using artificial intelligence in smart cities while keeping people accountable for public decisions.…

Fiddler AI Lists Seven Controls to Put in Place Before AI Agents Run Alone
Safety and ethicsNovember 4, 2026

Fiddler AI Lists Seven Controls to Put in Place Before AI Agents Run Alone

Fiddler AI has published a guide for regulated enterprises setting out seven practices to adopt before any AI agent operates without a human reviewing each step. The…