Why is it in the news?

• The European Union Parliament has passed the groundbreaking AI Act, marking the first major legislative effort globally to regulate Artificial Intelligence.

 More about the news

• The AI Act aims to safeguard fundamental rights, democracy, rule of law, and environmental sustainability, all while fostering innovation within the AI sector.
• It introduces a framework categorizing AI systems into four levels of risk:

1) Unacceptable Risk: Any AI system posing a threat to EU Fundamental Rights falls under this category and is prohibited.

2) High-Risk: AI systems with potential impacts on health, safety, and Fundamental Rights are subject to conformity assessments and ongoing monitoring.

3) Specific Transparency Risk: AI systems with risks related to manipulation or impersonation must fulfil information and transparency obligations.

4) Minimal Risk: Common AI systems such as spam filters are considered to have minimal risk and are not subject to specific regulations.

• General-purpose AI models with systemic risks must assess and mitigate these risks, report serious incidents, and undergo state-of-the-art tests.
• Real-time remote biometric identification in publicly accessible spaces, like facial recognition via CCTV, is prohibited with limited exceptions.
• Efforts to mitigate racial and gender biases in AI systems are mandated, requiring high-risk systems to be trained with adequately representative datasets to minimize biases.