EU AI Act delays, Anthropic soars, and India moves

AI regulation is getting rewritten just as the model race speeds up: Europe is loosening some deadlines, Anthropic is racing toward a trillion-dollar valuation, and India says its AI framework is almost ready. Here are the must-know moves. ^[3][8][10]

EU AI Act shift: negotiators reached a provisional Digital Omnibus deal on 7 May, pushing some high-risk deadlines back and adding new bans on AI-generated intimate imagery and CSAM. Big signal: Europe is simplifying, but not softening. ^[3]

Anthropic update: Financial Express says the company is near a $1 trillion valuation after a $65B round, while Opus 4.8 adds a 2.5x fast mode and new workflow controls. The takeaway? Enterprise AI is getting pricier and more powerful at once. ^[8]

India’s AI framework is closing in. Technosports says MeitY is finalizing a light-touch approach built around transparency, bias reduction, and data sovereignty, alongside the IndiaAI Mission’s ₹10,371.92 crore compute push. ^[10]

And in the US, Trump backed away from a planned AI safety review order after pressure from tech leaders, according to The Guardian. It is another reminder that AI policy is still being shaped as much by power as by technical risk. ^[15]