World’s first major act to regulate AI

The European Union (EU) parliament passed World’s first major act to regulate AI (Artificial Intelligence).

Key Highlights of the Act:

Fundamental Objectives:

• Protection of Fundamental Rights: The framework seeks to safeguard fundamental human rights, democracy, and the rule of law from the adverse impacts of AI.
• Environmental Sustainability: Ensuring AI applications do not harm the environment but instead contribute to its sustainability.
• Promotion of Innovation: Encouraging the development of AI technologies in a manner that is safe, ethical, and respects fundamental rights.

Risk Categories for AI Systems:

• Unacceptable Risk: AI systems posing threats to EU Fundamental Rights will face prohibition.
• High-Risk: Systems impacting health, safety, and Fundamental Rights must undergo a conformity assessment and continuous monitoring.
• Specific Transparency Risk: Systems that could manipulate or impersonate must adhere to stringent information and transparency obligations.
• Minimal Risk: Common AI systems like spam filters face no specific regulations, reflecting a balanced approach to innovation and safety.

Regulations for General-Purpose AI Models:

• Mandates include risk assessment, mitigation measures, reporting of serious incidents, and conducting advanced tests to address systemic risks.

Biometric Identification:

• The regulation generally prohibits real-time remote biometric identification (e.g., facial recognition) in public areas, with certain exceptions, prioritizing privacy and personal freedoms.

Addressing Biases:

• High-risk AI systems must utilize datasets that are sufficiently representative to minimize racial, gender, and other biases, promoting fairness and equity.

Ref:Source