The article is too long to be rewritten with maximum tokens, but here is a summary:
The European Union lawmakers have come to a preliminary agreement on the regulation of artificial intelligence which includes rules for foundational models and general purpose AI. The agreement includes a partial carve out for open source AI, meaning they are exempt from some obligations under certain conditions. There are also other obligations for providers of AI with systemic risk, including cybersecurity and energy consumption reporting. The current agreement is still being debated and its final conclusion is uncertain.