Context:
In continuation of pledges made during last year’s Bletchley Park AI Safety Summit, USA and UK signed an agreement on AI safety.
More on News
- The collaboration aims to create tests for advanced artificial intelligence (AI) models.
- The USA and UK both have agreed to share important information about AI model’s capabilities and the risks associated with it
- They will also work together on basic research about AI safety and security. This agreement is in effect immediately.
- The reason behind this cooperation is to find ways to safely use of AI as it’s spreading quickly.
- Although AI offers many benefits but it also brings risks like spreading fake news and affecting the fairness of elections.
Significance
- This is the first agreement of its kind worldwide to strive to maximize AI potential to solve big problems while minimizing the risks it brings.
- Both countries will work together to develop “robust” methods for evaluating the safety of AI tools, their systems, and agents.
- The US and UK AI Safety Institutes also plan to develop a common way to test AI safety and share their knowledge to handle risks effectively which will help other countries in shaping their own regulations.
- It also intended to tap into a collective pool of expertise by exploring personnel exchanges between the Institutes.
- In future both countries will develop similar partnerships with other countries to promote AI safety across the globe.
Challenges of AI regulation:
- The Red Queen problem highlights the subtle evolution of Artificial Intelligence, such as Generative AI and Reactive machines. Consequently, governments might lack the proficiency needed to adequately regulate and assess AI tools.
- As AI continues to advance rapidly there is a debate on what aspects to regulate and the methods to do so to ensure the safety of AI’s performance.
- Regulating too much could hinder innovation and limit AI’s potential benefits.
- Also, it’s hard to predict and address all the risks linked with AI, which means some risks (Deep fake, etc.) might go unnoticed.
What actions are India and the world taking regarding AI regulation?
- Recently, India’s IT Ministry issued an advisory to generative AI companies to get government approval before deploying untested systems.
- After facing criticism government withdrew this advisory and issued a new one without the requirement for approval.
- Last year the EU agreed on the AI Act with member states introducing rules for AI use within the EU including restrictions on its use by law enforcement.
- Consumers within EU countries now have the right to file complaints if they feel these rules are violated.
- The US White House also issued an Executive Order on AI that provides a model for other countries to follow in regulating AI.