The EU AI ACT came into force on 1st August 2024

Sources: EU, Deloitte, William Fry

It's been described by some as GDPR on steroids. If you think the Act doesn’t apply to you because you’re not based in the EU, think again

• The EU AI Act “The Act” is a risk-based approach to AI systems (from prohibition to gentle approach), while also recognising the power of LLMs

• Advances in technology and the growing role of AI require thoughtful inclusion strategies to ensure fair and equitable treatment of all individuals

• The main purpose of the EU AI Act is to mitigate risks and promote responsible AI use across sectors

•  The Act defines an "artificial intelligence system (AI system)" as: 

•  "machine-based system that is designed to operate with varying levels of autonomy and that can, for explicit or implicit objectives, generate outputs such as predictions, recommendations, or decisions that influence physical or virtual environments".

• The Act has been described by some as GDPR on steroids with some fines for non-compliance exceeding those for GDPR breaches. Like the GDPR, the Act is broad in scope and organisations will be expected to comply with both sets of regulations. 

• Given AI’s reliance on data (data), governance is key

• The EU believes the Act will foster public trust in AI tools while encouraging innovation and competitiveness in the EU market by creating a level playing field for developers & providers who intend to serve EU customers or process EU origin data. Such developers & providers will have to meet EU AI Act obligations whether they are based in Boston, Berlin or Bangalore

Leaders relying on AI tools to recruit, market, supply or do other business with people in the EU should familiarise themselves with the EU AI Act, which came into force on 1 August 2024 

• While compliance dates will be staggered, it’s important to note that rules on Prohibited AI Systems as well as AI Literacy Requirements come into effect as early as 2 February 2025

• Non-technical leaders who believe the EU AI Act doesn’t apply to them or their organisations need to think again including those in HR who use recruitment and/or performance tools, L&D (training tools), educational institutes & vocational training (admission & assessment tools), supply chain (I have written about AI & supply chain issues here , financial institutions (credit scoring systems), and more, who may be using AI tools classified as ‘High Risk’ due to their potential impact on health, safety & fundamental risks

• Non-compliance may, in some cases, result in fines of up to €15m or, up to 3% of total world-wide revenue for the preceding year, whichever is higher

In some cases, non-compliant AI systems can be taken off the EU market

Examples of the AI Act which Leaders need to be aware of include although not limited to:

1.  ARTICLE 4: ‘AI LITERACY’ - Organisations must ensure their staff possess adequate AI literacy (depending on their technical knowledge & experience), to enable them to make informed decisions about the AI tools they are using as well as repercussions should they change use case. Supplying incorrect, incomplete, or misleading information in response to a request can, in some instances, result in fines of up to €7.5m or, up to 1% of total world-wide revenue for the preceding year, whichever is higher. Key Date: 2 Feb 2025

2. ARTICLE 5: PROHIBITED AI SYSTEMS - AI systems that exploit vulnerabilities based on age, disability, or socio-economic status, social scoring AI systems, as well as emotion recognition in work places & educational institution unless used for safety (pilots for example) or medical reasons.  Non-compliance may, in some cases, result in fines of up to €35m or, up to 7% of total world-wide revenue for the preceding year, whichever is higher. Non-compliant AI systems can be taken off the EU market. Key Date: 2 Feb 2025

3. ARTICLE 6: CLASSIFICATION RULES FOR HIGH-RISK AI SYSTEMS - Includes AI systems used in: biometrics, educational institutions & vocational training, employment & work management (recruitment & performance AI tools), as well as AI tools used to access essential services including social services. Non-compliance may, in some cases, result in fines of up to €15m or, up to 3% of total world-wide revenue for the preceding year, whichever is higher. In addition, supplying incorrect, incomplete, or misleading information in response to a request can, in some instances, result in fines of up to €7.5m or, up to 1% of total world-wide revenue for the preceding year, whichever is higher. Key Dates: 2 August 2026 & 2 August 2027.

ARTICLE 14: HUMAN OVERSIGHT - Under Article 14, the Act calls on human oversight to minimise risks to health, safety and fundamental rights caused by automation bias.

ARTICLE 51:  GENERAL-PURPOSE AI MODELS & SYSTEMS WITH SYSTEMIC RISK - Using GPAIs and then tweaking/customising them for your business requirements will not exempt you from Union copyright & related rights. The EU AI Act regulations in this regard is to ensure there is a level playing field for AI developers/providers within the EU and that no third party provider has a competitive edge by applying lower copyright standards than these provided by the Union. Fines: up to 3% of total worldwide annual turnover from the previous financial year or €15 million, whichever is higher. Key Date: 2 August 2025 

Using GPAIs and then tweaking/customising them for your business requirements will not exempt you from Union copyright & related rights

So, whether you Build or Buy, the AI Act will apply to you if you employ people, have customers in, process EU origin data or, do business in the EU, irrespective of where your HQ is based.

 ******************************************

Subscribe to my TWO Newsletters and receive alerts as soon as new editions are published: (1) Inclusive Leadership in the era of AI  and/or (2) Leading with Emotional Intelligence (EQ) 

ABOUT ME - I'm a Consultant specialising in ‘Inclusive Leadership and AI’ advising C-Level executives, leaders, and those in succession, on navigating the dynamic intersection of leadership, technology and inclusivity. My 'Inclusive Leadership & AI' programme begins with fostering Emotional Intelligence (EQ) and Cultural Intelligence (CQ), recognising these as fundamental pillars in navigating the transformative and ever-changing environments shaped by emerging technologies including AI. I am deeply passionate about the transformative potential of AI - not just as a tool for speed, efficiency, or automating mundane tasks but also, as a tool for enhancing inclusion, self-improvement (by speeding up the learning process) and medical breakthroughs, to name but a few AI use cases. My focus and what I am looking to help build is Emotional Intelligence (EQ) in AI - utilising my 15+ years of experience in global recruitment, Executive Search & Coaching. I am also the co-founder of two tech startups: a digital recruitment platform; and a geocode system tailored for emerging markets.  If you wish to learn more feel free to reach out to me on LinkedIn.