EU AI Act enforcement has started: what businesses should do now

Enforcement does not mean every AI rule applies today

From 2 August 2026, the European Commission's AI Office and national authorities began enforcing applicable provisions of the EU AI Act. Article 50 transparency requirements also started to apply on that date.

For Irish businesses, this coincides with the AI Office of Ireland becoming operational as the country's central coordinating authority for implementation of the AI Act.

The AI Act still follows a phased timetable. Some obligations apply now, while parts of the high-risk regime have later application dates. The practical response is neither panic nor delay. It is to identify which rules apply to the AI your business actually uses.

Prohibited-practice rules are already applicable. General-purpose AI provider obligations have applied since August 2025. Article 50 transparency rules now apply. Following the July 2026 reforms, Article 4 requires providers and deployers to take measures supporting the development of AI literacy among staff and others using AI on their behalf. It does not require them to guarantee a specific level for every individual.

Complaints are now part of the enforcement picture

The Commission has established an AI Act Complaint Tool. Individuals and organisations can use it to report alleged infringements involving AI systems that fall within the AI Office's enforcement remit. It is not an anonymous reporting channel.

A separate secure AI Act Whistleblower Tool allows eligible people professionally connected to relevant AI providers or deployers to report suspected violations anonymously. National market-surveillance authorities handle much of the wider enforcement of AI systems.

Ireland uses a distributed regulatory structure. The AI Office coordinates implementation and supports cooperation between existing sectoral competent authorities. For businesses, that makes clear internal ownership, records and escalation routes increasingly important.

Five things to check now

If an authority asks how your business manages AI, evidence of what you have actually done is stronger than saying staff were told to 'use AI responsibly'.

• Know where AI is being used. Create a basic inventory covering approved software, AI features inside existing platforms and public tools such as ChatGPT used by employees.
• Identify higher-risk uses. Pay particular attention where AI influences recruitment, employees, customers or important decisions about individuals.
• Give employees practical AI rules. Staff should know what information can be entered into AI tools, when outputs require verification and when human review is necessary.
• Review customer-facing AI. Article 50 transparency rules apply to certain interactive AI systems and AI-generated or manipulated content. Relevant chatbots, for example, may need to make clear that the customer is interacting with AI.
• Keep evidence. Document your AI inventory, policies, staff guidance, vendor assessments and material decisions.