In world first Europe reaches deal on AI

Subject : Science and Tech

Section: Awareness in IT

Context: Europe agrees landmark AI regulation deal

More about the news:

The European Union (EU) reached a provisional deal on groundbreaking regulations for artificial intelligence (AI), addressing concerns such as the use of AI in biometric surveillance and the regulation of AI systems like ChatGPT.
The agreement requires foundation models and general-purpose AI systems to comply with transparency obligations, conduct evaluations, and report on various aspects.
Governments are restricted in using real-time biometric surveillance, and the agreement bans practices like cognitive behavioral manipulation and untargeted scrapping of facial images.
The legislation, once formally ratified, is expected to come into force in two years and could influence global AI governance standards.

Some details about the EU framework:

The European Union (EU) has reached a provisional deal on AI legislation, incorporating safeguards for consumer complaints, restrictions on facial recognition technology, and penalties for rule violations.
The legislation classifies AI applications into risk categories, with certain applications, such as mass-scale facial recognition, being largely banned.
High-risk applications, like AI in self-driving cars, will require certification and transparency.
Medium-risk applications, such as chatbots, will have documentation and transparency obligations.
The EU aims for this legal framework to support trustworthy AI development and provide a launchpad for EU startups and researchers in the global AI race.

What is the Legal Framework Related to AI in India:

In India, there are currently no specific legal regulations governing the use of deepfake technology. However, existing laws can be applied to address the misuse of this technology, covering aspects such as Copyright Violation, Defamation, and cybercrimes.
For instance, the Indian Penal Code, which addresses defamation, and the Information Technology Act of 2000, which pertains to sexually explicit material, could potentially be used to combat malicious deepfake usage.
The Representation of the People Act of 1951 contains provisions that prohibit the creation or dissemination of false or deceptive information about candidates or political parties during election periods.
Additionally, the Election Commission of India has established regulations requiring registered political parties and candidates to obtain prior approval for all political advertisements on electronic media.
Despite these measures, they may still be inadequate in fully addressing the multifaceted challenges arising from AI algorithms, including the potential risks associated with deepfake content.

What are the Recent Global Efforts to Regulate AI:

The world’s inaugural AI Safety Summit, hosted at Bletchley Park in the UK, saw 28 major nations, including the US, China, Japan, the UK, France, India, and the European Union, unite in signing a declaration emphasizing the necessity for global action to address the potential perils of AI.
The declaration underscores the recognition of significant risks stemming from potential deliberate misuse and unintended control challenges in advanced AI, particularly in domains such as cybersecurity, biotechnology, and the spread of disinformation.
In response to these concerns, the US President issued an executive order aiming to fortify defenses against AI-related threats and exercise regulatory oversight over safety standards applied by companies in the assessment of generative AI systems like ChatGPT and Google Bard.
During the G20 Leaders’ Summit held in New Delhi, the Indian Prime Minister advocated for the creation of a global framework governing the development of “ethical” AI tools.
This shift in New Delhi’s stance signifies a transition from a position of non-interference in AI regulation to a proactive approach, involving the formulation of regulations grounded in a “risk-based, user-harm” perspective.