The European Union has issued a formal request for Microsoft to provide detailed information regarding the risks associated with the generative AI features in its Bing search engine. This request, made under the Digital Services Act (DSA), highlights concerns over potential misuse of AI technologies, especially with the upcoming European Parliament elections scheduled for June 6-9, 2024.
Specific AI Features Under Investigation
The EU’s inquiry primarily focuses on two specific generative AI features in Bing. The first is the Copilot chatbot, powered by OpenAI’s GPT-4, which assists users by generating text-based responses. The second is the Image Creator by Designer, which enables users to create images from textual descriptions. EU officials are particularly concerned that these features could facilitate the creation and spread of deepfakes and other misleading content, potentially manipulating voter perceptions and undermining electoral integrity.
Risks Identified by the European Commission
The European Commission has identified several risks associated with these AI tools, including:
- Dissemination of Deepfakes: Generative AI could produce realistic but fake images and videos, which might be used to deceive the public.
- Automated Manipulation: There is a fear that AI could automate the spreading of misleading information, influencing public opinion and voter behavior.
- Hallucinations: AI systems sometimes generate false or misleading information, referred to as “hallucinations,” which can further complicate the information landscape.
Microsoft’s Response and Compliance Efforts
Microsoft has been given a deadline of May 27 to comply with the EU’s request. Failure to provide the necessary information could result in fines up to 1% of its total annual income, with additional penalties for providing incomplete or incorrect data. Microsoft has stated that it is fully cooperating with the European Commission and is committed to addressing the regulatory concerns raised. The company emphasizes its ongoing efforts to measure and mitigate potential risks associated with its AI technologies.
The Digital Services Act and Its Implications
The DSA, which came into effect in February 2024, aims to curb the online distribution of illegal content and misinformation. It imposes stricter regulations on large online platforms, like Bing, requiring them to implement robust measures against harmful content. This act is part of the EU’s broader strategy to ensure digital safety and uphold the integrity of democratic processes within its member states.
Broader Implications for AI and Big Tech
This scrutiny of Microsoft’s Bing AI is part of a larger trend where regulatory bodies globally are increasingly vigilant about the implications of AI technologies. The EU’s actions signify a growing emphasis on accountability and transparency from tech giants, particularly in how their innovations might impact society. As generative AI continues to evolve, regulatory frameworks will likely adapt to address new challenges and ensure these technologies are used responsibly.
The EU’s demand for information from Microsoft regarding Bing’s AI features underscores the significant regulatory focus on generative AI’s potential risks. With the deadline looming and substantial fines at stake, Microsoft’s response will be closely watched as a bellwether for future interactions between tech giants and regulatory bodies.
