Tech Giants Lobby Against EU’s AI Act

Tech giants Microsoft, Google and OpenAI have all visited European institutions to lobby against the upcoming EU AI Act. The Act is set to classify AI tools into high-risk and low-risk categories, with high-risk tools not being allowed in Europe. The companies are arguing that their tools should not be considered high-risk, as they are not intended for a specific use.
Sam Altman, CEO of OpenAI, has publicly advocated for some form of legislation, but has also argued that rules should be tailored to the company’s own products. OpenAI sent a white paper to the EU Commission and Council in September 2022, which argued that their tools are not high-risk.
The final draft of the AI Act, approved in June 2023, no longer states that general-purpose AI systems should be considered high-risk by definition. Instead, it requires transparency about the datasets used in the training of important models, such as the language model for ChatGPT.
OpenAI argued that their approach to mitigating risk arising from the general nature of their systems is industry leading. However, it is possible to circumvent the built-in security in OpenAI language models, as evidenced by Jailbreak Chat, which is full of prompts to abuse the language model.
Overall, the tech giants are attempting to lobby their way out of the AI Act, as they want to gain traction with the general public. They are arguing that their tools should not be considered high-risk, as they are not intended for a specific use. It remains to be seen whether their efforts will be successful.