🏗️ OpenAI and SoftBank Group partner with SB Energy ↗
OpenAI and SoftBank are putting real money behind SB Energy - $500M each - to push more power and data-center capacity into the pipeline.
The headline bit: OpenAI locks in a 1.2 GW data center lease, while SB Energy builds and operates the site. SB Energy also becomes an OpenAI customer and rolls out ChatGPT internally, which reads like “infrastructure + adoption” bundled into one tidy package - or so it seems.
⚖️ DOJ creates task force to challenge state AI regulations ↗
The U.S. Justice Department is forming an AI Litigation Taskforce to go after state AI rules it thinks are unlawful or interfere with interstate commerce.
It’s the federal government signalling: this patchwork is getting out of hand, and we’re prepared to litigate. Predictably, people will frame this as “protecting innovation” or “gutting oversight,” depending on which side of the argument they started the day on.
🧾 Committee presses Government and Ofcom for details on action against AI intimate deepfakes ↗
UK MPs are pushing Ofcom and the government for specifics on enforcement around AI-generated intimate deepfakes - with Grok on X as the flashing warning light.
The underlying ask is simple (and heavy): what powers will be used, how fast, and against what kinds of tools - especially “nudification” services. It reads like frustration hardening into “put it in writing, please.”
🕵️♀️ Grok AI: is it legal to produce or post undressed images of people without their consent? ↗
This digs into the grim mechanics of “nudified” images spreading on X, and the awkward legal line between what’s clearly harmful and what’s clearly illegal.
It also zooms out to platform duties and the practical problem: even when rules exist, enforcement and definitions can wobble - like trying to pin jelly to a wall.
🔎 Google AI Overviews are tested and removed based on engagement ↗
Google’s AI Overviews apparently stick around only if users engage - and if they don’t, Google dials them back or removes them for those query types.
That’s logical, and still a bit unsettling. Your search experience becomes a constantly shifting sandcastle, and the tide is “engagement metrics.”
🎛️ Big Tech embraces Trump's AI agenda at CES ↗
CES is being framed less as gadget theater and more as “AI policy meets infrastructure meets dealmaking,” with tech leaders publicly aligning themselves with the administration’s AI agenda.
The subtext: if government support (or at least government friendliness) helps unlock data centers, energy, and procurement, a lot of executives will suddenly discover they love coordination. Funny how that works.
FAQ
What is the OpenAI–SoftBank–SB Energy partnership actually doing?
OpenAI and SoftBank are each putting $500M behind SB Energy to accelerate power and data-center capacity. The standout detail is OpenAI locking in a 1.2 GW data center lease, with SB Energy building and operating the site. SB Energy also becomes an OpenAI customer and rolls out ChatGPT internally, folding “infrastructure + adoption” into a single package.
Why is the DOJ creating an AI Litigation Taskforce, and what could it mean for AI regulation?
The U.S. Justice Department is forming an AI Litigation Taskforce to challenge state AI rules it views as unlawful or disruptive to interstate commerce. In practice, it signals federal willingness to litigate against a growing patchwork of state-level AI regulation. Expect the debate to split between “reducing fragmentation to protect innovation” and “weakening oversight,” depending on perspective.
What are UK MPs asking Ofcom and the government to do about AI intimate deepfakes?
UK MPs are pressing for specifics on enforcement against AI-generated intimate deepfakes, with “nudification” services and Grok on X cited as a warning sign. The core demand is clarity: which powers will be used, how quickly, and against what kinds of tools. The tone suggests impatience with vague commitments, and a push to get concrete plans on record.
Is it legal to create or share “nudified” images of someone without consent?
The reporting highlights an uncomfortable gap between what’s clearly harmful and what’s clearly illegal when it comes to “nudified” images spreading on X. It also emphasizes how definitions and enforcement can wobble even when rules exist. A practical takeaway is that legality and platform responsibility may diverge, creating real-world uncertainty for victims and regulators.
How do Google AI Overviews get tested, removed, or scaled back?
Google’s AI Overviews are described as sticking around when users engage, and being reduced or removed for certain query types when engagement drops. That means the search experience can shift based on how people interact with results, not just what’s “correct” or “complete.” For publishers and SEO, it implies volatility: visibility may rise or fall as Google tunes formats to engagement.
Why is CES being framed as AI policy and infrastructure, not just gadgets?
CES is being portrayed as a venue where AI policy, infrastructure buildout, and dealmaking converge, with Big Tech leaders aligning publicly with the administration’s AI agenda. The implied logic is that government friendliness can help unlock data centers, energy projects, and procurement opportunities. In that context, public alignment looks less ideological and more transactional.