Bal mukund Shukla’s Post

View profile for Bal mukund Shukla

Head of Business Transformation & AI for Financial Services | Managing Director & Sr Partner | CXO Advisor. | FinTech & Cloud Transformation leader | Forbes Council Member

The EU AI Act, effective since August 1 2024, introduces a forward-looking, risk-based approach to AI regulation, emphasizing consumer rights and safety. Here's a breakdown: - **Minimal Risk:** AI systems like recommender systems and spam filters fall here. - **Specific Transparency Risk:** Transparency required for AI Chatbots and AI-generated content with labels such as biometric categorization, emotion recognition systems, Synthetic audio, video, text & images. - **High Risk:** Strict rules for "Customer decisioning applications" such as Lending and Recruiting. - **Unacceptable Risk:** Banned if it infringes on human rights or free will. The AI Act at EU will be implemented and enforced by the Commission’s AI Office and supported by European AI Board, a scientific panel, and an advisory forum. It sets a full applicability timeline by August 2, 2026, with phased provisions enforcement. Prohibitions for unacceptable risk will be enforced after 6 months, governance rules for so-called General-Purpose AI models after 12 months, and AI system regulations will apply after 36 months. The AI Pact aids in compliance transition, urging early adherence to key obligations for global AI developers. Non-compliance can lead to fines up to 7% of global turnover for banned AI applications, 3% for other violations, and 1.5% for supplying incorrect information. #AI #Regulation #AIPact

Arvind Venkatesan

Capital Markets & Wealth Management | Industry Solutions & Consulting | Building Client Relationships | Delivering Business Transformation |

7mo

This is a great first step and needs to be enforced globally. With Gen AI bringing greater onus on Privacy risk, regulations need to cover that aspect too. Also AI regulations should not be looked in isolation but in conjunction with other related areas like cybercrime.

Like
Reply
Jayesh Kumar

Trusted partner and growth Navigator

7mo

Well summarised Bal! It is a well thought out regulation which I am sure will encourage responsible AI in every facet of its usage.

Like
Reply
See more comments

To view or add a comment, sign in

Explore topics