๐๐จ๐ฉ ๐๐ ๐๐ก๐ข๐ง๐ ๐ฌ ๐ญ๐จ ๐๐ง๐จ๐ฐ ๐ข๐ง ๐๐ก๐ ๐๐ซ๐ญ๐ข๐๐ข๐๐ข๐๐ฅ ๐๐ง๐ญ๐๐ฅ๐ฅ๐ข๐ ๐๐ง๐๐ (๐๐ญ๐ก๐ข๐๐ฌ ๐๐ง๐ ๐๐๐๐จ๐ฎ๐ง๐ญ๐๐๐ข๐ฅ๐ข๐ญ๐ฒ) ๐๐ข๐ฅ๐ฅ, ๐๐๐๐. It will have an impact on DPDPA as well.
It is a proposed Indian legislative framework designed to regulate the development and deployment of automated systems
- Creation of an Ethics Committee: The Bill establishes a dedicated Ethics Committee for Artificial Intelligence to oversee AI deployment, consisting of experts in ethics, technology, law, data science, and human rights.
- Broad Definition of AI: AI is defined as computer systems or applications capable of tasks requiring human intelligence, specifically including decision-making, language processing, and visual perception.
- Strict Surveillance Controls: The use of AI for surveillance is restricted to lawful purposes only and requires prior approval from the Ethics Committee.
- Anti-Discrimination in Critical Sectors: AI systems used in law enforcement, financial credit, and employment are prohibited from discriminating based on race, religion, or gender.
- Mandatory Transparency: Developers must disclose the intended purpose, limitations, data sources, and training methodologies of their AI systems.
- Right to an Explanation: Developers are required to disclose the reasons for any decisions made by AI systems that impact individuals.
- Compulsory Bias Audits: There is a legal mandate for developers to conduct regular audits to identify and mitigate systematic errors (algorithmic bias).
- Withdrawal of Biased Systems: If an AI system exhibits significant bias, it must be withdrawn from use until corrective measures are implemented.
- Grievance Redressal: Any individual or group affected by AI-related harm or misuse has the right to file formal complaints with the Committee.
- Hefty Penalties: Violations can result in fines up to five crore rupees, suspension of licenses, and potential criminal liability for repeat offenders.
๐๐ฆ๐ฉ๐๐๐ญ ๐จ๐ง ๐๐๐๐๐
The AI Bill is designed to function alongside existing laws, at par DPDPA,2023, rather than replace them. Specifically:
- Additive Compliance: The Bill states its provisions are “in addition to and not in derogation of” any other law in force. This means businesses mu st comply with both the AI Bill and the DPDPA simultaneously.
- Layered Regulation: While the DPDPA focuses on how personal data is collected and processed, the AI Bill regulates the algorithmic logic applied to that data. For example, while DPDPA ensures data is processed with consent, the AI Bill ensures the result of that processing is not biased or discriminatory.
- Enhanced Disclosure: The requirement to disclose “data sources” under the AI Bill adds a specific layer of transparency to the data processing obligations already established under the DPDPA.
This blog is an academic initiative brought to you by theย Data Privacyย Pro team ofย AMLEGALS. Subscribe โ Stay updated, Stay compliant.