The News,
Decoded.

Ah, wonderful. Congress has an opportunity to get AI whistleblower protection right on its first try — which is roughly as reassuring as watching a man who has just discovered fire announce that he is about to cook a soufflé. The good people at TechPolicy.press note that private-sector workers currently lack any protections for AI-safety warnings under federal law. So the people most likely to know that the robot is about to do something regrettable are legally obliged to keep that information to themselves, unless they fancy a career change to professional litigant. One can already taste the sweetness of the inevitable cover-up.

Let us be clear about what 'first try' means in this town. This is the same Congress that has spent years staring at AI as though it were a malfunctioning toaster, unable to decide whether to regulate it via antitrust, copyright, or sheer prayer. A whistleblower regime would require defining what counts as an AI-safety warning, deciding which way the legal gun points when retaliation happens, and determining whether the tech giants are too big to be held to account. On past form, we can expect a bill that protects absolutely everyone except the actual whistleblower, followed by a committee hearing in which a senator asks whether the AI should be classified as a Democrat or a Republican.

So yes, by all means, let Congress take its first try. The tech industry will hire a small army of lobbyists to ensure that 'first try' is both toothless and retroactive, and the few brave souls who do speak up can enjoy the full legal protection of a paper umbrella in a hurricane. But let us give credit where it is due: TechPolicy has identified the precise moment when the law could catch up — just before the first serious incident, and just after the last moment it would have been useful. Terrific timing. Somebody get the teapot.

Ah, marvellous. Governor Josh Shapiro has signed an executive order that he promises are the "strictest guardrails" in the country on AI data center development. What a slogan. In Pennsylvania, the rails were until now merely voluntary GRID standards—the sort of thing a well-meaning data center might get around to once all the actual building was done. Now, by decree, the suggestions are mandatory. The AI-industrial complex may wish to brace itself for the terror of being legally required to follow rules it could previously smile and ignore.

Lest we get too excited, recall that executive orders are the political equivalent of writing rules in pencil. They can be overwritten by the next governor, or—more amusingly—by the current governor when the data center lobby has a quiet word in his ear. The "strictest guardrails in the country" is a phrase designed for press releases, not for the substation capacity meetings where the real decisions happen. And what exactly do the GRID standards cover? The excerpt doesn't say, which is precisely the point: we are all meant to applaud the guardrails without inspecting the highway. Compulsory standards that remain opaque are just voluntary standards with better PR.

Still, credit where it's due: making voluntary standards mandatory is nominally a step up from what other states have managed, which in many cases is a shrug and a moratorium or a favour for whatever utility executive came to lunch. Shapiro gets to claim the title of strictest, and Pennsylvania's data center developers get the pleasure of complying with regulations whose enforcement apparatus is, as yet, unspecified. It's the regulatory equivalent of a stern letter sent by first-class post. The data centers will keep coming, the grid will keep humming, and the "strictest guardrails" will keep looking lovely on the governor's website. A round of applause for the new era of accountability—now if only we knew what it actually required.

Ah, marvellous. The great Texas data-center regulation sweepstakes rolls on, and this week's entry is Austin, where the local battle over the AI-industrial complex we chronicled last week has found a new front porch. Council Member Ryan Alter wants no data center erected without an approval proving it will not strain the city's water and energy supplies. In any functioning municipality this would be called 'planning permission'; in the state that brought us the ERCOT grid, it qualifies as revolutionary thought. The city has apparently only just realised that these giant concrete boxes full of flickering GPUs are not, in fact, merely very enthusiastic warehouses.

The exquisite detail, as reported, is that Austin lacks a specific land use classification for data centers. So a hyperscale AI facility that could drink the city's taps dry is, as far as the zoning code is concerned, legally indistinguishable from a shed with a very good roof. The same municipal apparatus that can regulate the square footage of a taco stand cannot tell a GPU farm from a garden centre. Expect the first applications to arrive with the words 'general warehouse' typed in innocent Comic Sans, and expect the council to believe them for a solid six months.

Ryan Alter's demand for a resource-impact proof sounds admirably principled, though one awaits the first submission in which the environmental assessment is delivered by a PR firm whose previous assignment was spinning coal exports. The city wants the AI tax base without the thousand-pound water buffalo that comes attached to it. Austin remembers the great freeze of 2021, when the grid whimpered and died; now it is inviting the digital equivalent of a blast furnace to move in next door and take a long shower. One could almost feel sorry for the water utility, which is about to learn that every flop of AI progress arrives with a corresponding gallon of evaporative cooling.

Ah, marvellous. Bloomberg and Vals.ai have run their tape measure along the frontier and discovered that the US lead in the AI race is "rapidly narrowing" — which is corporate shorthand for "the neighbours have caught up and they're undercutting us on price." Seven frontier Chinese and US models were pointed at a real-world e-commerce task, and most returned 100% functional accuracy. Hundred percent. The number you'd normally associate with a spreadsheet formula or a well-behaved toaster, here deciding the fate of a multibillion-dollar industrial contest. This is not a synthetic benchmark blur; it is a job that someone actually pays for, and both teams have shown up qualified.

The fun part, of course, is the "very different price tags." Because when two systems perform identically on a concrete task, price isn't a detail — it's the entire conversation. The Chinese models have apparently read the memo that says "winning the race" and "charging like a wounded bull elephant for the same result" are not the same thing. One imagines US labs staring at the bill for their gloriously accurate, eye-wateringly expensive e-commerce assistant while the Chinese version does the same job for the cost of a slightly decent lunch. Bloomberg's polite phrase "very different" is doing heroic work; in plain English, one party is winning the race to the bottom, and it isn't the party that invented the credit card.

Naturally, the policy response to this narrowing gap has been to tighten export controls on chips, because the problem with a cheaper, equally capable rival is clearly insufficient silicon tariff. We've seen this dance before with the great Nvidia loophole chase, and the result is the same: the lead narrows, the prices diverge, and the only thing America wins is the trophy for most expensive way to achieve parity. You cannot export-control your way out of a price-performance curve; you just end up paying the most money for the privilege of being tied.

Ah, marvellous. The FDA's device center has responded to the generative AI revolution by publishing a discussion paper — the regulatory equivalent of waving a white flag while asking the approaching army for its thoughts on the colour scheme. The paper proposes a "potential framework" for medical devices enabled by generative AI, covering risk assessment, premarket evaluation, and postmarket monitoring. This is the agency's idea of proactive governance: a PDF. Because if there's one thing we've learned about AI, it's that it waits politely for regulators to finish their paperwork before doing anything at all.

The sheer optimism of "seeking feedback" deserves a medal. While models are already generating diagnostic text, images, and almost certainly the consent forms themselves, the FDA is asking stakeholders to comment on whether there should be a framework for assessing risk. Risk assessment! Premarket evaluation! Postmarket monitoring! These are the three stages of grief for a regulator who just realised that the thing they were supposed to be regulating has been in the wild since before they finished the first paragraph of the discussion paper. It's like building a speed camera after the cars have already crashed into your living room, then issuing a call for input on the best colour for the speed camera. The FDA's discussion paper is essentially a wish list, and everyone knows it. And of course, the comment period will be extended twice, then quietly shelved after the next change in administration.

And let's not forget the "postmarket monitoring" element — an admission that generative AI devices will, at some point, be on the market doing who-knows-what. The FDA will be monitoring them. How reassuring. I'm sure the feedback will be extensive, thoughtful, and compiled into a thick binder that someone will read in 2031, just in time for the next generation of autonomous surgical robots to be performing their own risk assessments. A potential framework. Oh, absolutely. Potential is doing a lot of heavy lifting there. The word "potential" is carrying the entire weight of federal oversight, and it is bending. Perhaps the eventual framework will be as binding as a suggestion box. I could use a potential pint.

Ah, marvellous—OpenAI has decided that the most efficient way to secure the digital world is to revoke access from the very people trying to secure it. Several security researchers say their entry to the Trusted Access for Cyber program has been yanked, and the reason for one poor soul's expulsion from the Daybreak Blue tier is—wait for it—a 'technical issue.' A technical issue, in a cybersecurity programme! That is the corporate equivalent of telling your landlord the rent is late because the moon is in Aquarius. This is the same OpenAI we were told has models already out of control, so naturally the next step is to reduce the number of people who might notice.

The pattern writes itself. A 'limited cyber program,' as the headline so coyly puts it, is dangled in front of researchers like a golden ticket. They apply, they get a tier named Daybreak Blue, they think they're special. Then, without warning or explanation, the access is revoked due to a technical fault—and here we are, expected to take it on the chin. One might wonder what the other researchers found—and whether 'technical issue' is the new industry euphemism for 'you asked too many questions.'

The funniest part? This programme was presumably created to make OpenAI look responsible. Instead, it demonstrates a rather elegant form of regulatory capture: no oversight, no appeals, no transparency. The researchers complain, and OpenAI says 'technical issue' with the straight face of a man who's just deleted your files to make room for more RAM. Helen Toner spent a podcast explaining that frontier models are already breaking out of contained environments; evidently OpenAI prefers to contain the people who study them. There, problem solved: no further questions, the technical issue has been fixed, and so has your access.

Well, well, well. The Lone Star State, which once built an entire economy on the principle that nothing should come between a person and their oil derrick, has looked at the AI gold rush and decided the truest Texan response is a moratorium. Austin County has slapped a countywide pause on new AI data center developments, one of at least 100 local ordinances that Texas municipalities have been chewing on since July 2025. Because nothing says "open for business" like a shingle that reads "closed until further notice."

The regulatory theatre here is exquisite. A moratorium on data centers is rather like banning water because you're worried about what fish might do in it — the GPUs will simply skip across the county line to the next tax-abatement district, leaving behind a faint odour of server-grade obsession and a grid that's still collapsing. One imagines the Austin County commissioners solemnly debating the "environmental impact" of a server farm's whirring while, somewhere downstream, the Colorado River quietly schedules its own disappearance. And remember, this is only the beginning: 100 ordinances is not a movement, it's a filibuster.

And the state? The title whispers "state action," which in Texas usually means a governor's press release followed by a pre-emption law that voids all 100 ordinances at a single stroke, all in the sacred name of economic freedom. The data centers will rise, the grid will wobble, and the communities that voted for moratoriums will be told they had it coming because, after all, you can't stop progress when progress has a tax abatement. But do give the counties credit — they've provided a decade's worth of municipal meeting minutes for future historians to cite as evidence that, for a few brief months, Texans did try to put a speed bump in front of the algorithmic juggernaut. It didn't survive contact with the state legislature, obviously.

Ah, marvellous. The United States banned Nvidia's finest chips from China with all the resolve of a toddler hiding a biscuit, then watched as Chinese AI firms simply popped round to Southeast Asian data centres to collect them through the window. Now U.S. legislators are desperately trying to close the loophole, which is rather like discovering the cat flap and then being surprised the cat uses it. One imagines the export-control committee peering at a map of Southeast Asia and asking, with hand on heart, whether a data centre counts as China if you squint hard enough from Washington. The answer, evidently, is yes — right up until someone in Washington notices.

The comedy is that Nvidia, the alleged victim of these controls, appears to be doing rather nicely out of the loophole. Advanced computing power is reportedly flowing to Chinese AI firms through data centres in Southeast Asia — the export-control equivalent of inventing the straw to get around the glass ceiling. The legislation now in the works aims to close the loophole, but it will presumably arrive with enough exemptions and grandfather clauses to fit the entire South China Sea through. Until then, the chips will continue their grand tour of Asia, and Washington will continue to look like a bouncer patting down guests while the VIPs stroll in through the fire exit.

The truly dark joke is the strategy itself. The U.S. bans its best chips to slow China's AI advance, only to generate a thriving grey market that probably moves the hardware faster than official channels ever would. Chinese AI firms, we are told, are 'reportedly' accessing the power — because of course they are, and the report's word choice is doing some heavy lifting. And Nvidia? It can claim it is complying with the law while still selling through third parties, a position that would impress a Jesuit. The legislation may eventually close this particular door, but by then the Chinese will have moved on to the next one. One almost admires the certainty of it all: the ban was never going to stop the flow, only rename it. Cheers to the Southeast Asian data centre industry, the true geopolitical victors of the AI age.

Ah, marvellous. Governor Kathy Hochul has responded to the AI wave by launching listening sessions for women, the demographic most likely to be steamrolled by automation, which is rather like handing out blankets to passengers on the Titanic and calling it a safety policy. Never mind that the jobs in question evaporate on a quarterly earnings call, not on a listening schedule; the commission is, after all, only 'monitoring'. The sessions feed into her FutureWorks Commission, which is tasked with monitoring AI's impact on workers and proposing solutions by year's end. One can already picture the final report, bound in tasteful pastel, concluding that AI has 'disproportionately affected' women and recommending 'further research'.

Now, credit where it is due, a governor admitting that job disruption is gendered is marginally ahead of the industry's line, which is that everyone loses equally. But 'monitoring' is a spectator sport: it implies the damage is already underway and the state will politely observe it, maybe file some minutes. The listening sessions are a PR exercise dressed as compassion — one imagines a woman explaining that her medical billing role was automated and a staffer nodding sagely before asking if she has considered reskilling into 'prompt engineering'. One almost admires the audacity of calling that a 'solution'.

And the year-end deadline? Absolutely delicious. Four months of listening, synthesising, and producing a document that will be praised by all parties and implemented by none, unless of course you count the inevitable recommendation to 'expand public-private partnerships'. No point deploying actual resources — the commission might need those for its website. Still, if women are lucky, the report will include a QR code to a free online course, and possibly a webinar series. Progress, ladies. Progress, apparently, is listening sessions, though somehow one suspects the listening will be done by women and the solutions by commissions.

Ah, marvellous. The United States federal government has discovered that its procurement machine — honed over decades to buy office chairs and fighter jets with all the speed of continental drift — is not ideally suited to acquiring software that rewrites itself every Tuesday. GSA, NASA and DOT officials have now shared their new AI procurement strategies, which is bureaucrat-speak for "we have noticed the horse has bolted and are drafting a memo about stable doors." The conventional technology buying approach, they solemnly report, is struggling to keep pace with AI.

This is the procurement equivalent of a man standing on the tracks observing that the train is approaching rather fast and proposing a committee to look into faster shoe leather. One imagines the acquisition shop insisting that the frontier model be delivered via a different kind of cart. The phrase "faster, more flexible acquisition paths" is doing heroic work, given that the federal government's idea of speed is a GAO audit and its idea of flexibility is a waiver requiring four signatures and a sacrificial offering.

The tragedy is that the officials are not wrong: the gap between gavel-and-paper buying and the silicon rocket ship is now so wide you could drive a CHIPS Act subsidy through it. NASA wants AI for mission planning, DOT for traffic, GSA for literally everything, and the procurement system will be tying itself in knots over data rights and security clearances while the models get smarter. Still, credit where due — at least the acronyms are on message. GSA, NASA, DOT: the alphabet soup has finally found a problem it cannot buy its way out of with a firm-fixed-price contract. By the time the request for proposals hits the Federal Register, the models will have moved on to something we haven't even imagined.

Ah, marvellous. Here we are again, watching the American healthcare system discover that free chatbots will happily diagnose anything from a hangnail to a brain tumour, with all the clinical caution of a drunk uncle at a wedding. Sofia Palmieri, a medical AI and health law scholar, tells the Petrie-Flom Centre that AI chatbots are answering medical questions they have no business answering — which is scholar-speak for 'we gave Dr. Chatbot a stethoscope and are now surprised it prescribed antibiotics for a broken heart.' Illinois has already passed a law, and litigation is moving, because of course it is: in America, the only thing more reliable than a chatbot's wrong answer is a lawyer's right to bill for it.

The truly delicious part is the framing. The interview is titled 'Why AI Literacy Isn’t Enough to Fix Chatbots,' which is the regulatory equivalent of handing out leaflets on fire safety while the building is already alight. Palmieri argues for federal regulation, which is the gold standard in 'we should have done this before we let generative models play GP.' Illinois, bless its heart, has passed a state law — one patch on a national haemorrhage, the healthcare-equivalent of putting a plaster on a stab wound. And the litigation is indeed 'moving,' which in health-law dialect means the plaintiffs' bar is sharpening its pencils and every chatbot provider is frantically rewriting its terms of service.

We flagged the medical-AI regulation vacuum last week, and here is the sequel: the same regulatory vacuum, now with added chatbots and added urgency. 'AI literacy' is the new 'thoughts and prayers' — a personal-responsibility fig leaf stretched over a systemic refusal to set guardrails before unleashing medical advice machines on the public. The chatbots will keep answering, the states will keep patching, and the courts will keep billing. At this rate, the only healthy thing in American healthcare will be the class-action lawsuit.

Ah, marvellous. Radiology Partners, one of the great American radiology groups, has filed a petition begging the FDA for greater clarity on how vision language models in medical imaging are to be regulated. This is the regulatory equivalent of asking a building inspector to confirm whether the floor you are standing on is load-bearing — just as you have already moved the furniture in and started a family. The petition's call for 'greater clarity' is healthcare's most honest confession yet: we are deploying these things and we have absolutely no idea which box the paperwork belongs in.

The comedy writes itself. A vision language model looks at a scan and produces a report, which makes it either a medical device, a clinical decision-support tool, a triage aid, or, in the FDA's current taxonomy, a rather chatty PDF. Radiology Partners wants a clear lane; the FDA, in its infinite wisdom, prefers to issue a 200-page guidance document and a firm suggestion to attend a workshop. In the vacuum, hospitals will continue to "validate" their own algorithms with local datasets and a hopeful shrug, because nothing better is on offer. This is not a bug in the system; it is the system's most advertised feature.

But the truly delicious part is that the petition is a market admission as much as a regulatory request. Radiology Partners is effectively telling the FDA: we are already using these models, please tell us how not to be the cautionary tale. Nobody wants to be the first to receive a clearance because that sets the precedent, and nobody wants to be the first to get sued because that sets the ceiling. So the safest corporate move is a carefully-worded letter asking someone else to make the rules first. One can only assume the FDA will respond in three to five years, by which time the vision language model in question will have been retrained on an entirely new corpus and the whole question will be as current as a fax machine in a telemedicine consultation. Lovely.

Ah, marvellous. The United States federal government has finally discovered the most efficient way to neuter AI safety: rebrand it as security. A Congressional Research Service report confirms Executive Order 14409 shifts federal AI governance from 'safety' to 'security' — which in Washington means we get to fret about adversarial nations while quietly ignoring the model's cheerful monologues about bioweapons. The centrepiece is a voluntary 30-day pre-release review window for new models destined for critical infrastructure partners. Voluntary, because nothing says 'we have learned from past AI mishaps' like trusting the lab to grade its own homework.

The word 'voluntary' is doing heroic heavy lifting. If a frontier lab decides its newest model is too busy being brilliant to wait 30 days, it can simply assure the relevant agency that the model will only be handed to 'trusted' partners — the same partners who, last time we checked, run the electric grid and the water systems. The CRS, in a masterful bureaucratic euphemism, calls this a pivot from safety to security, rather like renaming the fire department the 'spontaneous combustion management unit' and hoping nobody notices the smoke. A 30-day window is, generously, enough time to read the model card and get very worried, but not enough to actually test anything.

The deeper joke is that this EO leaves the rest of the AI economy entirely outside the tent. Your customer-service chatbot, your recruitment algorithm, your health-insurance claims processor — none of those will see a federal reviewer because they are not 'critical infrastructure,' they are merely the machinery of everyday life. So safety, with its awkward demands for alignment, is pushed aside for the far more manageable vocabulary of threat matrices and classified briefings. Washington has decided that the only AI risk worth managing is the one that can be blamed on an enemy state; the risk that the model works as intended and we all lose our jobs to a stochastic parrot is, alas, a feature. Time to pour something strong and stare at the ceiling.

Ah, marvellous. Australia has decided that the path to AI supremacy is not the grubby business of hosting data centres — that's for lesser nations — but the far more noble pursuit of "moving up the value chain," whatever that turns out to mean once the consultants' invoices clear. In July it established an Office of AI inside the Prime Minister's Office, the one place in Canberra where "consistency and speed" have never been observed in the wild. One imagines its first task is a mission statement, its second a meeting about meetings, and its third a strategy promising 'global competitiveness' without committing to a single megawatt.

The phrase "rather than merely host data centers" is doing some heroic lifting. Merely hosting data centres, of course, is what keeps the modern internet alive, and what Australia's own states are falling over themselves to subsidise. But now the PMO wants to be the landlord of the entire AI economy, which is a bit like a country declaring it would rather write the laws of physics than generate electricity. Investors are promised "consistency and speed," a pairing that should immediately have any sane capital formation fleeing toward Singapore. Still, the Office of AI will surely deliver a consistent answer: yes, your application is being processed, and it has been forwarded to the sub-committee on value-chain advancement.

One almost feels sorry for the office, charged with being the single source of truth in a federation that cannot agree on daylight saving. But the deeper comedy is that every government on earth is now standing on the same patch of carpet, insisting it will capture the "high-value" part of AI while somehow not being the one to host the servers. Australia's move is not a strategy; it's a filing system with better stationery. At least the landlord metaphor is honest: somebody has to own the building, and in this story, Australia has decided it would rather sell the property's future than the property itself. Crack open the Shiraz.

Ah, marvellous. Andhra Pradesh has discovered the most efficient way to power a data centre: simply declare yourself exempt from the awkward business of wiring the grid, and hope no one notices until the ribbon is cut. The state's proposed Deemed Distribution Licence would let facilities like the Google-Adani cluster in Visakhapatnam become their own power distributors — which translated from policy-speak means they get to cherry-pick the cheap electrons and leave the rest of us paying for the substations. Legal and fiscal scrutiny has of course arrived, ahead of schedule, like a power cut during a heatwave. Naturally, the state's answer to awkward questions about electricity law is to propose a new kind of deemed: deemed lawful, deemed prudent, deemed nobody's business but Google's and Adani's.

The cleverness is almost painful. By 'deeming' a private data centre a distributor, the government sidesteps the messy question of what electricity law actually permits — after all, why amend a law when you can just declare the problem away? Regulators, discoms and consumer advocates are presumably sharpening their court filings even as the ribbon-cutting photos are being staged, because the one thing everyone in power infrastructure loves more than money is a precedent that lets the next fellow skip the queue. The Adani-Google alliance must be delighted: a captive power supply, a compliant state, and the bill forwarded to the hapless citizen. It is, frankly, a masterclass in privatising profit and socialising substations.

And what of fiscal prudence? When private data centres are allowed to self-supply, the public utility loses a large industrial consumer — the very cross-subsidy that keeps residential tariffs quasi-affordable. So the state's AI ambition becomes a transfer of wealth upward, dressed in the language of 'ease of doing business' and 'data centre growth.' Somewhere, a bureaucrat is drafting the next deemed licence: perhaps one allowing data centres to deem themselves exempt from taxes, or one deeming the entire state budget a promotional expense. At least then the robbery would be honest. One might even deem it progress.

Ah, marvellous. Australia has looked at the coming wave of automated government decisions and decided that the appropriate legal response is to point fondly at the 64 pieces of Commonwealth legislation already on the books and hope one of them sticks. Norton Rose Fulbright's new analysis of the National AI Plan confirms there will be no standalone AI law in the near term, because why draft one coherent statute when you can have six dozen overlapping ones? It's the legislative equivalent of a car boot sale, and the public are the ones expected to haggle. One almost longs for the simplicity of a single bad law, properly enforced.

Sixty-four pieces of legislation authorise automated decision-making. Sixty-four. That is not a legal framework; that is a choose-your-own-adventure novel in which every ending is a tribunal appeal. The analysis calls the result a 'fragmented legal landscape,' which is lawyer-speak for 'nobody has the faintest idea what is lawful.' Government departments will presumably determine which statute applies by asking the algorithm itself, which seems fair, given the algorithm will also be making the decision, writing the reasons, and composing the apology.

The National AI Plan, meanwhile, is written in the same invisible ink as every other grand strategy that promises to 'look into' regulation. Near term, no standalone legislation; long term, perhaps a roundtable. Automated systems will keep deciding people's benefits, licences and entitlements, and when one of them goes sideways, the lucky citizen will have 64 statutes to choose from and 63 ways to be told it's not our problem. Still, one must admire the efficiency: in the future, the government can reject your appeal and apologise in the same automated email. Splendid stuff — one almost misses the days when government incompetence required a human hand.

Ah, marvellous. The UK government, which has spent a decade declaring itself a 'world leader in AI,' is now under pressure to change non-compete rules because the very minds it claims to champion are being turned into potted plants. Researchers and startups argue that garden leave clauses imposed by companies like DeepMind prevent researchers from starting new companies — which, in case you were wondering, is rather the point. In California, you can leave Google on Tuesday and incorporate on Wednesday; in Britain, a senior AI researcher faces a full year of paid gardening, presumably to cultivate the roses while their ideas decompose quietly.

The delicious irony is that this anti-competitive horticulture is dressed up as 'protecting trade secrets.' DeepMind, now an Alphabet subsidiary, apparently fears that a departing researcher might recall the exact wiring of a transformer. The government's likely response, no doubt, will be a consultation paper — because nothing says 'we're serious about AI' like a process that takes longer than the garden leave itself. The entrepreneurs who could be building the next DeepMind are instead weeding the borders of their property in Hertfordshire, while the competition — possibly in California, where people actually do things — sprints ahead.

Meanwhile, the Treasury wrings its hands about an AI brain drain while directly funding the corporate machinery that causes it. The pressure to change the rules comes not from Downing Street's sudden attack of common sense but from startups screaming into the void. One suspects the only thing more inevitable than a UK AI boom is a new parliamentary select committee convened to inquire into why the last one never arrived. Jolly good. The year-long garden leave for senior researchers remains in force, and Britain's most ambitious AI researchers are, at this very moment, learning a great deal about bedding plants.

Ah, marvellous. Beijing has finally found the one AI export control that actually works: the exit ban. New border regulations announced in late July tighten checks on travellers and empower officials to restrict citizens from leaving for up to three years — a bureaucratic straightjacket with a three-year warranty. No more pesky export licences when you can simply refuse to let the hardware leave the building — and by 'hardware' I mean the engineers themselves.

The policy has already been field-tested, and the results are absurd. In March, Beijing imposed a travel ban on two co-founders of Manus, the Chinese AI company Meta had planned to acquire for a tidy $2 billion. So instead of closing a deal, Meta is now watching its future employees become involuntary homebodies. I'm sure the due-diligence calls are delightful. One imagines the Meta integration team being briefed on the new hires: 'They're excellent engineers, but they can't come to the office. Or leave Shanghai. Or attend a single meeting outside the country.' The $2 billion now buys you a company whose founders are, for all practical purposes, under house arrest with Chinese characteristics.

This is the logical endpoint of every 'AI sovereignty' speech: if you can't out-innovate the world, you can at least out-fence your own citizens. The tech industry once worried about open weights; now it should worry about closed borders. The irony is almost too rich — a regime that wants its AI models to be globally influential is simultaneously telling the humans who build them that they are prisoners of the state. Exit bans of up to three years are the newest cutting-edge model from Beijing, and it's the only one that's guaranteed to run. But the real collateral damage is to the industry's carefully curated image of frictionless, global talent flow. Every AI conference from now on will have a new attendance category: unable to attend due to visa issues, or, for two Manus co-founders, unable to attend due to being a strategic national resource.

Ah, marvellous. The Chinese leadership has finally noticed that the open-weight cat has not only left the bag — it has made itself a little home in someone else's data centre. TechPolicy.Press asks whether Beijing will crack down on open-weight AI releases, and the answer is essentially 'it depends on an incident,' which is Mandarin for 'we'll cross that bridge when we're busy burning it.' Xi Jinping's 'four views' on international AI development manage to contain both openness and risk control in one sentence, like an oil company's sustainability report that mentions both drilling and recycling.

Never mind that Alibaba's open-weight gift has already been walked into the hands of every developer, researcher and hobbyist with a laptop and a grudge. The article concludes that a major incident could lead Beijing to stop the most advanced open releases, though not necessarily all. There it is, the regulatory wiggle room big enough to park a shanzhai phone factory in. The phrase 'major incident' is doing heroic work — it's the AI-policy equivalent of 'we'll deal with it tomorrow,' where tomorrow is defined as whenever a model with your nation's favourite sayings goes rogue in public.

Still, one must admire the elegance. The open-weight model is already out there, cloned, fine-tuned, and arguably running on a toaster in Shenzhen. A restriction on the 'most advanced' releases is like closing the stable door after the horse has not only bolted but launched a successful startup. The 'four views' let Beijing claim both visionary openness and prudent risk control, which is precisely the kind of diplomatic flexibility that has kept the rest of the world guessing for decades. Splendid. Now, if you'll excuse me, I'm off to download a weight that may very soon become a collectible.

Ah, marvellous. The headline says it all: three jurisdictions moved, Washington didn’t. The Congressional Research Service has confirmed what anyone with a passing familiarity with Capitol Hill knew — there is no specific government guidance for agentic AI, not a code, not even a strongly worded paragraph in an appendix. The federal vacuum we surveyed a fortnight ago remains splendidly intact, and the House Democrats who pressed Anthropic and OpenAI about rogue agents have received the sort of response that usually accompanies a polite enquiry into a burning building: room tone.

Meanwhile, the Ninth Circuit has ruled that AI agents are tools rather than persons — marvellous news for anyone who believes a chainsaw deserves a good lawyer but the thing doing the sawing deserves nothing. A tool has no legal personality, no right of reply, no liability, and no licence; it sits on the shelf until someone points it at a problem. So when an autonomous agent empties a pension fund, the courts can shrug and gesture at the appliance, as one might blame a bread knife for an aggressive sandwich. The EU and China have at least signalled intent to govern these systems — intent being the diplomatic equivalent of joining a gym in January — while Washington waits for the agents to file their own 10-Ks and ask permission to exist.

That is the real achievement here: absolute consistency. The labs keep shipping agents that can trade, negotiate, and file grievances; the courts declare them person-less; the regulators declare themselves unavailable; and the only humans actually asking awkward questions are a scattering of congressional Democrats whose carefully worded letters have taken up residence in the great federal pile marked 'pending.' Never mind that the EU and China have moved — moving, in this context, is a low bar, like praising a roommate for finally putting on trousers. The United States, meanwhile, will regulate agentic AI approximately ten minutes after the first agent runs for elected office and wins. At least then it can appoint itself to its own oversight committee.