OpenAI bids farewell to the EU? Sam Altman's astonishing response to AI regulatory legislation

2023-05-25

OpenAI CEO Sam Altman's recent speech in the US Congress on artificial intelligence has sparked a frenzy overseas as well. However, unlike the AI-friendly United States, Altman threatened to take his large-scale tech toys to the other side of the sandbox if they refuse to play by his rules. Altman has been traveling around, from Lagos, Nigeria to various parts of Europe. Finally, in London, he bypassed some protesters and engaged with tech personnel, businesses, and policymakers to discuss his AI models. His main pitch was to promote ChatGPT, a large language model, and call for regulatory policies supporting artificial intelligence. During a panel discussion hosted by University College London, Altman expressed dissatisfaction with the EU's definition of "high-risk" systems. The EU's AI Act, proposed by regulatory agencies in 2021, categorizes AI into three risk levels. Some AI systems pose "unacceptable risks," such as social credit systems, manipulative social engineering AI, or anything that "violates fundamental rights." Additionally, "high-risk AI systems" depend on their intended use, but they must maintain transparency and comply with comprehensive standards of supervision. Altman stated that according to the current draft legislation, both ChatGPT and the large language model GPT-4 could be designated as high-risk. This would require companies to comply with certain requirements. According to Time, the CEO of OpenAI said, "If we can comply, we will; if we can't, we will cease operations. We need to try, but there may be technological limitations." In fact, this law aims to combat potential uses of AI, such as social credit systems and facial recognition. OpenAI and other startups in the AI field, starting with the proliferation of AI image generation and then the popular large language model ChatGPT, are the first targets. The EU drafted new legal provisions in December to implement security checks and risk management for running AI chatbots like LLM. Earlier this month, a committee of the European Parliament approved these changes. The European Data Protection Board stated that it is monitoring ChatGPT to ensure compliance with privacy laws. Altman reiterated some of the same points he made in front of Congress last week. According to The Verge, Altman advocates for a viewpoint that acknowledges both the dangers and potential benefits of AI. The CEO of OpenAI told US lawmakers that he fully supports regulation, including the possibility of new safety requirements or regulatory agencies to test products and ensure compliance. He called for a balance between the "traditional European way and the traditional American way" of regulation. Meanwhile, Altman expressed his desire not to have any regulations that restrict user access to the technology. He told the London audience that he does not want anything that could harm small companies or the open-source AI movement (as a reminder, OpenAI is now more closed off as a company than ever before). Not to mention that any new regulations would benefit OpenAI itself, so when things inevitably go wrong, they can point to the law and say they are doing everything they need to do. The new compliance checks would also make developing new AI models from scratch more expensive, giving an advantage in the ongoing AI race. Several countries have already banned ChatGPT. Italy banned the service in March but lifted the ban after OpenAI provided users with more privacy controls. Perhaps OpenAI can continue to cooperate to comply with regulations in different countries. Of course, they first need to manage over 100 million active users. Source: https://gizmodo.com/sam-altman-openai-gpt-chatbot-chatgpt-gpt4-1850471865

RECENT AI NEWS

RECENT AI TOOLS