According to the AI Act, European Parliament’s priority is to make sure that AI systems used in the EU are safe, transparent, traceable, non-discriminatory, and environmentally friendly. The European AI Act classifies AI according to its risk:
- Unacceptable risk is prohibited. Therefore, the following types of models should not be used:
- a. Subliminal, manipulative, or deceptive AI
- b. Systems that exploit vulnerabilities related to age, disability, or socio-economic circumstances to distort behaviour, causing significant harm.
- c. Biometric categorisation systems inferring sensitive attributes (race, political opinions, trade union membership, religious or philosophical beliefs, sex life, or sexual orientation), except labelling or filtering of lawfully acquired biometric datasets or when law enforcement categorises biometric data.
- d. Social scoring systems.
- e. Systems assessing risk of an individual committing criminal offenses.
- f. Compiling facial recognition databases from the internet or CCTV footage.
- g. Inferring emotions in workplaces or educational institutions.
- h. ‘Real-time’ remote biometric identification (RBI) in publicly accessible spaces for law enforcement.
- High-risk AI systems: they are regulated, and the AI Act focuses mostly on these.
- Limited risk AI systems: they are subject to lighter transparency obligations. This means, developers and deployers should make sure that end-users are aware that they are interacting with AI (thus making it clear that there is an AI model behind chatbots and deepfakes).
- Minimal risk AI models are unregulated: these include most AI applications that were available on the EU single market at the moment of starting AI Act in 2021. For instance, AI-enabled video games and spam filters.
Clearly, this scenario is changing with generative AI, which raises the level of risk of AI models, making them mostly high-risk. Even if now AI regulations are advancing, we think AI systems should be overseen by people, rather than by automation, to prevent harmful outcomes. This includes making sure we are able to generate trustworthy AI systems that are fair by design and are explainable and clear to decision makers. Legislation will clearly not fast-track adoption of responsible AI but organisations are the ones who need to share experiences and solutions to show what “good practice” looks like. Boards need to embrace Corporate Digital Responsibility to assess digital impacts of products/services on all stakeholders by examining societal, economic, technological & environmental impacts. Therefore, we are supporting companies in their role to ensure the technology is not deployed in “negative use cases” that could harm society and generating AI models that are transparent and accountable across industries.
AI explainability is becoming a regulatory requirement, not just a governance best practice
While transparency has always been a key principle of trustworthy AI, the European AI Act significantly strengthens the need for explainability across the AI lifecycle. For high-risk AI systems, providers must ensure sufficient transparency so that deployers can understand, interpret, and appropriately use system outputs. This includes providing documentation on system capabilities, limitations, expected performance, risk scenarios, and, where applicable, information that helps explain how outcomes are generated.
The AI Act also links explainability to human oversight requirements. Human decision-makers must be able to effectively monitor AI systems, challenge outputs when necessary, and intervene to prevent harmful outcomes. In practice, this means organisations will increasingly need explainability techniques, model documentation, audit trails, and monitoring tools that make AI decisions understandable to both technical and business stakeholders.
Beyond high-risk applications, new transparency obligations for generative AI systems require users to be informed when interacting with AI and when content has been generated or manipulated by AI. As these requirements take effect, organisations will need stronger governance frameworks that demonstrate not only compliance, but also that AI systems can be trusted, interpreted, and held accountable. Here you can see the guidelines from the European Commission on transparency obligations for providers and deployers of AI systems.
The AI Act’s implementation is increasing the demand for AI explainability
The AI Act has now moved from legislation to implementation:
- 2 February 2025: prohibitions on unacceptable-risk AI and AI literacy obligations became applicable.
- 2 August 2025: obligations for providers of General-Purpose AI (GPAI) models came into force.
- 2 August 2026: transparency obligations and enforcement powers for important parts of the Act started applying.
This means, in essence:
- Transparency obligations are now coming into force, making explainability more than an ethical aspiration.
- Human oversight requirements depend on understanding AI outputs and limitations, increasing demand for explainable systems.
- The GPAI Code of Practice and related guidance emphasize documentation and transparency, accelerating adoption of governance and explainability tools.
→ Check our Trustworthy AI solution