The AI Act has been introduced in Europe to address the risks associated with AI systems and establish clear boundaries for their use. It sets obligations for both users and developers, proposes a governance structure at national and bloc levels, and outlines an assessment guideline. However, exemptions are made for open-source projects and scenarios where AI innovation supports small and medium enterprises (SMEs) to avoid excessive regulatory oversight.

One significant objective of the AI Act is to prevent AI systems from generating illegal content. While well-known generative AI products like OpenAI’s Dall-E and ChatGPT, Microsoft’s Bing Chat, and Google’s Bard have safety measures in place, there are publicly accessible AI tools that lack such filters.

These tools can be misused to create manipulated media, including explicit deepfakes. In fact, the FBI recently issued a warning regarding the increase in deepfake crimes. Additionally, AI systems may generate false “facts” due to inherent problems like “hallucinations.” It’s worth mentioning that Europe is not the only region actively working on AI regulation, and it will still take months for the AI Act to be legally enforced.

In April, the Commerce Department sought public input to shape AI policy recommendations, particularly with regards to necessary safety measures at the federal level. Likewise, China’s internet regulator released a comprehensive proposal on AI product regulation, aligning with the country’s strict censorship laws.