New AI Rules of European Union

The EU’s recently passed EU AI Act, which will be rolled out in phases over the next two years, has ignited a battle over data transparency.

Key Highlights On New AI Rules of European Union

• Template: The EU’s AI Office plans to release a template for organisations to provide “detailed summaries” of the content/ data used by organisations who deploy general-purpose AI models, such as ChatGPT, to train their AI models following a consultation with stakeholders.
• Objective: To ensure an appropriate balance between the legitimate need to protect trade secrets and, on the other hand, the protection of privacy and creators’ rights over their content, including copyright holders’ rights to exercise their rights under Union law.
• Datasets as Trade Secret: AI companies are highly resistant to revealing what their models have been trained on, describing the information as a trade secret that would give competitors an unfair advantage if made public.
• Ethically Sourced Content: Technology companies have signed a flurry of content-licensing deals with media outlets and websites. 
  • Example: OpenAI signed deals with the Financial Times and The Atlantic, while Google struck deals with NewsCorp (NWSA.O), opening a new tab social media site Reddit.

Enroll now for UPSC Online Course

Need for Transparency Rules

• Breach of copyright: In the past few years, a few big tech companies, including Google, OpenAI, and Stability AI, have faced lawsuits from creators claiming their content was improperly used to train their models.
• Ensure fair remuneration: There is a growing call for tech companies to pay fairly, the copyrights holders for data used by them to train their AI models.
• Right of consent and privacy: Every creator should have the right to know if their work, songs, voice, art, or science was used in training the algorithm.
  • Example: OpenAI has faced backlash for featuring an AI-generated voice described as “eerily similar” to her own by actress Scarlett Johansson in a public demonstration of the newest version of ChatGPT.

Challenges

The Data transparency rules have the potential to hinder the European AI startups’ competitiveness.

• Affect Europe’s AI ecosystem: The rules can result in the shift of the AI ecosystem from Europe to other countries and Europe can be  reduced into becoming a consumer of American and Chinese products.
• Bad Regulation: Pre Mature regulation of emerging Technologies like AI can hinder innovation in the sector running the risk of regulating technologies that haven’t been mastered, or regulating them badly because it has not been mastered.
• Widespread Implication: The rules will have big implications for smaller AI startups and big tech companies like Google (GOOGL.O), and Meta (META.O), which have put the technology at the centre of their future operations.

Way Forward

• Awareness of Rights: People should have the awareness that they are communicating  or interacting  with an AI system as well as duly informing users of the capabilities and limitations of that AI system and affected persons about their rights.
• Traceability: AI systems shall be developed and used in a way that allows appropriate traceability and explainability  in case of any wrong and harmful information. 
• Transparency: The technical infrastructure of AI systems should be transparent in their functioning so that users can understand the process and logic behind the decisions made. 
  • This includes providing an explanation of how an AI system arrived at its decisions, as well as information on the data used to train the system and the accuracy of the system.
• Risk based disclosure Framework: To address the industry concerns and to drive innovation the Rules should not be blindly applied to all the AI systems, rather needs to be rationalised as per the risk it poses.
  • Example: For deep fake-generating systems, it must be disclosed that the content has been artificially generated or manipulated in a clearly visible manner.

Check Out UPSC CSE Books From PW Store