The AI Act Is Already Here. Do You Know Which Rules Apply to Your Organisation?
With another compliance deadline approaching on 2 December 2026, understanding which AI Act obligations apply — and to whom — is increasingly important.
Artificial intelligence is becoming part of everyday business operations, from drafting documents and analysing information to supporting recruitment, customer service and decision-making.
But as organisations adopt AI tools, an important question arises: What are their legal responsibilities?
The EU Artificial Intelligence Act (Regulation (EU) 2024/1689) establishes a regulatory framework for AI systems. Its requirements are being introduced progressively, with different obligations applying depending on the type of system, its intended purpose and the organisation's role.
Key AI Act Deadlines:
What Businesses Need to Know
What changes on 2 December 2026?
Two important developments take effect on 2 December 2026.
1. Transparency requirements for AI-generated content
Under Article 111(4), providers of AI systems generating synthetic audio, images, video or text that were placed on the market before 2 August 2026 must comply with the marking and detection requirements of Article 50(2).
These requirements aim to ensure that AI-generated content can be identified as artificially generated or manipulated, subject to the Regulation's conditions and exceptions.
2. Additional prohibited AI practices
New prohibitions concerning certain AI systems used to generate or manipulate prohibited intimate or child sexual abuse material also become applicable.
Importantly, this is not a general compliance deadline for every organisation using AI. The applicable obligations depend on the AI system, its intended purpose and the organisation's role.
What changes on 2 December 2027?
From 2 December 2027, the AI Act's requirements for high-risk AI systems classified under Article 6(2) and Annex III will become applicable. These include certain AI systems used in areas such as:
-
Employment and recruitment: AI systems used to screen job applications, evaluate candidates or make decisions affecting employment.
-
Education: AI systems used to assess learning outcomes or determine access to educational opportunities.
-
Essential services: Certain AI systems used to assess eligibility for essential public benefits or services.
-
Critical infrastructure: AI systems used as safety components in the management and operation of specified critical infrastructure.
-
Law enforcement, migration and justice: Certain AI systems used in these sensitive areas.
Depending on their role, organisations providing or deploying these systems may need to implement measures relating to risk management, data governance, documentation, transparency, human oversight and monitoring.
Not every AI system used in these sectors is automatically high-risk. Classification depends on its intended purpose and the specific conditions established by the Regulation.
What changes on 2 August 2028?
From 2 August 2028, the AI Act's requirements for certain high-risk AI systems covered by Article 6(1) and Annex I become applicable, subject to relevant transitional provisions.
These include AI systems used as safety components of products, or forming part of products, covered by specified EU product safety legislation, where the conditions for high-risk classification are met. Examples may include certain AI-enabled medical devices, lifts and other regulated products.
Depending on their role, organisations developing, supplying or deploying these systems may be subject to requirements concerning risk management, technical documentation, conformity assessments, human oversight and ongoing monitoring.
Importantly, not all high-risk AI systems follow the same compliance timetable. Organisations should determine how their systems are classified, identify the obligations applicable to their activities and prepare accordingly.
What does this mean for businesses?
The AI Act does not impose the same obligations on every organisation using artificial intelligence.
The applicable requirements depend on the organisation's role, the AI system's intended purpose, its risk classification and the relevant implementation deadlines.
For example, a business using an AI writing assistant for routine administrative tasks may have different responsibilities from an organisation developing or deploying an AI system to evaluate job applicants or make employment-related decisions.
The key is understanding which rules apply to your organisation, and when.
Businesses should therefore identify the AI systems they use or provide, assess their intended purposes and determine whether specific requirements apply.
They should also remember that the AI Act operates alongside existing legislation, including the GDPR whenever personal data is processed.
Compliance begins with understanding the technology, its use and the responsibilities that follow.
Where should organisations start?
Rather than introducing new policies or procedures without first understanding their needs, organisations should begin by assessing how artificial intelligence is being used within their operations.
A few practical questions can help:
-
AI use: Which AI tools and systems are currently being used, and for what purposes?
-
Roles and responsibilities: Are we developing, providing or deploying AI systems, and who is responsible for their oversight?
-
Risk classification: Could any of these systems fall within a prohibited or high-risk category?
-
Legal obligations: Which AI Act requirements apply to our activities, and when do they become applicable?
-
Existing safeguards: Are appropriate controls, internal guidance and human oversight in place?
-
AI literacy: Do employees have sufficient knowledge to use AI tools appropriately and understand their capabilities and limitations?
The answers can help organisations identify compliance gaps, prioritise necessary actions and determine where additional safeguards, training or specialist guidance may be needed.
The goal is not to introduce unnecessary complexity, but to ensure that AI is used responsibly, with appropriate safeguards and a clear understanding of the obligations involved.
Beyond the deadline
The AI Act is not simply about meeting regulatory deadlines. It also raises broader questions about how organisations introduce and use artificial intelligence, manage risks and maintain appropriate oversight.
As AI becomes increasingly integrated into everyday operations, organisations need to understand not only what the law requires, but also how those requirements translate into practice.
Effective compliance is not about creating more procedures than necessary.
It is about putting the right measures in place, at the right time, for the right reasons.
Add comment
Comments