Enterprises may be racing to adopt AI, but the challenges are coming into sharper focus as risks and real-world consequences surface.There are some weird, wild and very personal questions that businesses have to ask themselves — for instance, should customers be allowed to have deep or romantic relationships with company chatbots?

China announced a ban on Nvidia’s custom RTX Pro 6000D chips for the Chinese market, marking a turning point in the race for AI dominance: Faced with being cut off or restricted from U.S. advanced chip supplies, Beijing would rather turn to its domestic alternatives, so it is forcing tech giants like Alibaba and ByteDance to pivot from reliance on U.S. semiconductors.

If the AI boom were a theme park, Nvidia has made a second deal for fast-pass access to compute capacity — first through a $1.5 billion agreement with AI cloud provider Lambda for Nvidia to lease back up to 18,000 of its own GPUs, and now a deal with CoreWeave for up to $6.3 billion worth of access to unused data center capacity, putting Nvidia at the front of the line.

Talk about large-scale: Yesterday, Alphabet became the fourth company ever to reach a $3 trillion market value, joining Nvidia, Microsoft, and Apple. It’s no coincidence that so much of their valuations are driven by investors’ enthusiasm for all things AI.

The pace of investment in AI infrastructure is accelerating worldwide, with billions pouring into new ventures, massive cloud agreements, and record levels of data center construction.

The not-usually-so-contentious trade and tech relationship between the U.S. and South Korea is top of mind this week, with the convergence of several developments.

AI infrastructure continues to drive eye-popping numbers: Right now, the headliner is Microsoft’s deal for GPUs from Nebius (which was spun out of Russian internet company Yandex two years ago).