The News,
Decoded.

So Google has unleashed Gemma 4 onto the world, complete with 'speculative decoding' — a technique so named because it involves the model speculating wildly about what tokens come next, then backpedalling when it gets it wrong. And we're meant to be impressed by this? The claim is 'up to 3x faster' on consumer GPUs, which in Google-speak probably means 'on an RTX 5090 that costs more than a used car, under ideal conditions, with a single prompt.' Still, three times the speed means three times the hallucinations per second, which is technically progress.

Let's not forget this is an *open* model, meaning anyone with a reasonably powerful graphics card can now generate slop at triple the pace. Google frames this as democratisation; I frame it as a free-for-all where every basement-dwelling hobbyist can now flood the internet with AI-generated nonsense at unprecedented velocity. The 'Multi-Token Prediction drafters' sound like something from a Terry Pratchett novel, but they're real, and they're here to make your Twitter feed three times more polluted.

And the best part? Google's own track record with open models is spotty — remember when they promised transparency then quietly pivoted? I'm sure this time will be different. After all, what could possibly go wrong when you give the internet a faster way to generate content that's *already* outpacing our ability to detect it? Oh, and it runs on consumer GPUs, so your next electric bill will be a lovely surprise. Cheers.

Ah, Microsoft — the company that brought us the Blue Screen of Death and a decade of patch-Tuesday purgatory — has decided that *frontier AI models* should be subject to government cyber testing. Because nothing says 'trust us' like the same firm that treated cybersecurity as an afterthought for thirty years now volunteering to have NIST poke at its shiny new brain. The gall is almost admirable. Microsoft's agreement with NIST's CAISI sounds lovely, but let's not forget that this is a company whose flagship OS was once so porous that a single email attachment could cascade a worm through the Pentagon. Now they want to test frontier models for *cybersecurity risks*? How about we start with the risks of handing the keys to a monopoly that has historically treated security as a cost centre rather than a feature.

The timing is convenient: the White House is considering an executive order requiring pre-approval for frontier models, and here comes Microsoft, sweeping in with a self-own dressed as a virtue. 'Yes, govern us, please!' they cry, while quietly knowing that their relationship with NIST gives them a front-row seat to shape the very tests that will — or won't — catch catastrophic flaws. It's the regulatory equivalent of a fox volunteering for a henhouse inspection, insisting on writing the checklist.

Of course, the real joke is that government cyber testing is only as good as the testers. NIST, for all its competence, moves at the speed of a congressional subcommittee on a Friday afternoon. Frontier models evolve faster than bureaucrats can update a protocol. By the time CAISI certifies a model as 'secure', it'll be as relevant as a Windows 95 firewall. Microsoft knows this. We know this. But hey, performative responsibility is the currency of the day, and the company is minting it. Cue the ribbon-cutting ceremony for the golden cage.

The White House is considering an executive order to require government pre-approval for releasing frontier AI models. Because nothing says 'innovation' like having to ask your ham-fisted uncle for permission before you open the toy. The tech policy and industry circles are 'freaking out,' which is exactly the reaction one hopes for from a regulation that reads like it was designed by a committee of people who've only read about AI in Tom Clancy novels.

They wanted clear, consistent rules to build around; they got a bureaucratic bottleneck that would make the DMV look like a speed-dating event. One imagines the order will specify a 'frontier AI model' as any model that could plausibly write a tweet about the cancellation of the F-35, or something equally tangible.

In practice, pre-approval means that the most advanced AI systems will either be delayed until Congress figures out its own login credentials, or they'll be outsourced to jurisdictions with less paternalistic impulses. Either way, the American AI sector gets to spend the next few years filling out forms instead of building things. But hey — at least someone in government is taking an interest. That's the first step, right? Right?

SAP is forking out €1 billion ($1.16 billion) to acquire Prior Labs and turn it into an 'AI lab for structured data' over four years. Because naturally, what the AI boom has been crying out for is a massive ERP conglomerate to buy a startup and spend four years slowly bureaucratising its spark into enterprise middleware. The German government must be thrilled — finally, a tech champion that keeps AI safely trapped in spreadsheets.

'The focus on structured data' — that's the new euphemism for 'we promise this won't do anything embarrassing like write poetry or question authority.' Prior Labs presumably had some clever algorithms for making sense of SAP's legacy databases, which is a bit like buying a Ferrari to tow a caravan. The four-year timeline is particularly delicious: whatever AI strategy they have now will be thoroughly obsolete by year two, but the office building will be lovely.

And of course, this follows the grand German tradition of overpaying for startups that might have been cheaper to build in-house (RIP Wirecard's ambition, but hello tax write-offs). The real winners here are the venture capitalists who got a 1.16-billion-dollar exit for a company most people hadn't heard of until SAP's PR machine woke up. Cheers to four years of 'synergising structured data' — may the quarterly reports be ever in your favour.

Ah, another day, another $27 million thrown at the 'interface layer between humans and AI agents' — because what we really need is a snazzy name for the chatbot wrapper that will, eventually, send an email to the wrong client. CopilotKit, an open-source darling, has just bagged a Series A to build infrastructure for 'agentic' applications. At least they're honest about the buzzword count.

The enterprise product they're launching sounds suspiciously like 'we put a chat window on your website and called it an agent.' But hey, it's open-source, so your company's data can leak with community oversight. The funding round is modest by AI standards — barely enough for a year's compute on a frontier model — but enough to keep the hype train chugging. Meanwhile, actual safety research remains underfunded.

The logic seems to be: if we can't make agents reliable, we can at least make them easy to deploy. Because what the world needs is more half-baked autonomous code that 'surprises' its creators. The interface layer between humans and AI agents is, apparently, venture capital. Cheers to progress — we'll be debugging the inevitable mess. (For a reminder of how well open-source agentic code works, see our earlier piece on Google DeepMind unionising over military AI deals.)

The UK government has announced a task force to 'examine whether and how' AI-generated content should be labelled. Because we've only had two years of deepfakes, election interference, and granny-scam voice clones to work out that maybe we should. The task force will produce an interim report by autumn. That's a solid three-to-four months for a group of civil servants to decide if water is wet. No word on what happens if the report concludes 'yes, we probably should' — presumably another task force to decide on the font.

This is the same government that has been 'looking at' AI regulation since 2023, producing white papers, consultations, and now a task force that will examine labelling. The phrase 'examine whether and how' is doing heroic work here — it lets ministers claim action while punting any actual decision to the next parliament, or the one after that. The interim report in autumn? That's conveniently close to the next party conference season. Coincidence? Of course not — nothing in British politics is accidental, except the meaningful stuff.

Meanwhile, AI-generated content is already flooding every platform. Deepfakes of politicians, fake product endorsements, and synthetic news articles are running riot. The task force's brief — to 'examine whether' labelling is needed — suggests we're still at the nodding-along stage while the problem accelerates. In the time it takes to read this summary, a hundred AI-generated Facebook posts will have been shared by retirees who think they're real. But don't worry: a report is coming. In autumn. Possibly with recommendations. Possibly.

Ah, how quickly the corporate conscience moves. A new Littler survey of 300 US business leaders reveals that 84% now expect AI policy changes to affect their operations—double last year's 42%. Meanwhile, the proportion fretting about DEI policy has plummeted from 84% to a mere 38%. It seems diversity initiatives have become the forgotten middle child in the boardroom, elbowed aside by the shinier, more terrifying prospect of actually having to regulate the workplace's newest overlord.

The speed of this pivot is breathtaking—and not in a good way. Last year, DEI was the hot topic, 84% worried. This year, it's barely half that, because apparently the only thing that unites corporate America is the shared panic over a technology they've already deployed without much thought. Littler, the law firm that conducted the survey, must be rubbing its hands at the billing opportunities: one day diversity compliance, the next day AI accountability, and the client base hasn't even blinked.

So farewell, DEI consultants; hello, AI ethics advisors (with the same PowerPoint, just swapped in 'bias detection' for 'unconscious bias training'). The real joke is that both concerns are rooted in the same fear: that the powerful will do something reckless, and the underlings will pay the price. But at least AI feels more futuristic, doesn't it? To the 38% still worried about DEI: your moment has passed. The 84% have spoken, and they're now busy drafting policies that will be, one suspects, just as toothless as the ones they've just abandoned. Cheers.

Ah, Connecticut — the Land of Steady Habits, and apparently the Land of Steady Committees. State Senate Bill 5 has cleared the statehouse and is toddling off to the governor's desk, promising to create 'AI oversight committees' and 'workforce development programs.' Because nothing says 'we're on top of this existential technological revolution' quite like forming a working group. One can almost hear the meeting minutes multiplying.

The timing is, as ever, impeccable. Frontier labs are shipping multimodal models that can write code, generate video, and probably draft their own regulatory exemptions, while Connecticut's finest are asking 'who should sit on the committee?' The bill is a masterclass in legislative impotence: by the time these committees actually produce a report, the AI in question will be asking to join the committee itself. Still, it's a start — if by 'start' you mean 'a very slow, deliberative, and almost certainly obsolete gesture in the direction of concern.'

And let's not overlook the workforce development angle — a classic sop. The bill's language about retraining workers for the AI economy is the legislative equivalent of a comforting pat on the head before the robot takes your job. Connecticut's workers can rest easy knowing that as their positions are automated, they'll have access to programs designed by committees that met twice and forgot to set an agenda. SB 5: because when in doubt, form a committee.

Ah, the classic 'we have concerns about ethics' gambit — now with added trade union muscle. Employees at Google DeepMind in London have voted to unionize, with the explicit aim of blocking the lab's technology from reaching the US and Israeli militaries. Because nothing says 'ethical AI' like needing a collective bargaining agreement to enforce it. Google's 'Don't Be Evil' motto apparently now requires a legally recognised worker representative to remind the board what that even means.

The union's primary objective is to halt contracts that would see DeepMind's AI used in military applications. The US and Israeli militaries, it turns out, are not quite what Google had in mind when they painted those 'AI for social good' murals in the office. It's a bold move — workers using their only real leverage (their labour) to draw a line in the sand. Whether that line holds against the weight of government contracts and shareholder returns remains to be seen.

Of course, the irony is delicious: a company built on the premise of superintelligence now finds its own employees distrusting the direction of that intelligence. The union vote is a cry of 'stop' from the very people building the future. But as anyone who has watched a corporate ethics board dissolve into irrelevance knows, a union contract is only as strong as the willingness to strike. And striking against a military contract isn't exactly a walk in the park. Still, raise a glass to the DeepMind workers — they've at least given the board something to sweat about over their oat milk lattes.

Ah, the classic 'ask for forgiveness not permission' approach to enterprise AI. ISACA's latest survey drops a delightful little data point: 90% of digital trust professionals believe employees in their organisation are already neck-deep in AI tools, yet fewer than half have any safety or security policies in place. That's like letting teenagers borrow the family car but refusing to teach them how to brake.

Look, we've all seen this play out. Marketing is feeding customer data into ChatGPT for campaign copy, HR is using AI to screen CVs with hallucination-prone models, and legal is still drafting a 'generative AI acceptable use policy' that will be obsolete by the time it's approved. ISACA's 'digital trust professionals' — a job title that sounds like it was invented by a committee trying to sound reassuring — are essentially the canaries in the coalmine, and the canary is already face-down in the slurry.

Of course, the real joke is that this gap between adoption and policy is not a bug but a feature. Speed over safety, innovation over caution — that's the Silicon Valley mantra that's now being swallowed wholesale by every HR and IT department. After all, if your competitor is already using AI to write code and you're still debating whether to allow it, you're already behind. What could possibly go wrong? Besides everything.

Ah, Common Sense Media is launching the Youth AI Safety Institute, an 'independent' lab to crash-test AI tools for risks to children. Because nothing says 'trust us, we're independent' quite like being 'industry-backed.' That's like hiring the fox to audit the henhouse—and paying him in chicken feed. The institute will presumably test whether chatbots can be tricked into giving a teenager instructions on how to build a pipe bomb, then issue a helpful 'amber alert' sticker. And if the industry sponsors happen to get a flattering safety rating? Pure coincidence, we're sure.

But let's not be too cynical: this is the same Common Sense Media that has spent years telling parents which movies have too much swearing, now pivoting to tell them which AI models might groom their children. A noble pivot, to be sure. Yet the track record of self-regulatory initiatives in tech is... how shall we put this? Littered with well-meaning announcements that quietly fizzle out once the press cycle ends. The Children's Online Privacy Protection Act has been around for decades, and we still have apps scraping toddler data like it's going out of style.

So the Youth AI Safety Institute will run its 'independent crash tests,' probably award some kind of 'Safer with Kids' badge, and the industry will point to it as proof that everything is fine—right up until the first major incident involving a teenager and a persuasive, unhinged chatbot. Then maybe, just maybe, we'll get actual legislation. But until then, we have 'independent' labs funded by the very people who need regulating. Cheers to performative progress.

Brian Armstrong's latest missive to Coinbase staff is a masterclass in corporate euphemism: 14% of the workforce out, and the company will no longer employ 'pure managers.' Because nothing says 'we value our people' like eliminating the layer of humans who supposedly coordinate them. This is the trend we noted last week with Coinbase's earlier cuts, except now they've given it a fancy label: the 'pure manager' is extinct, with AI now handling performance reviews and resource allocation. How efficient. How soulless.

The Business Insider piece calls middle managers the most layoff-prone job in tech, and honestly, who's surprised? We've spent years building systems that automate decision-making, and now we're shocked that the humans in the decision-making chain are redundant. The logic is impeccable: if an AI can schedule your stand-ups and compile your weekly reports, why pay someone a six-figure salary to do it with more meetings? Armstrong's letter didn't mention the word 'AI' explicitly in the excerpt, but the subtext is clear: the algorithm doesn't need a middleman to deliver bad news.

Of course, what this really means is that the people who could have caught subtle team issues, mentored juniors, or pushed back on stupid directives are being replaced by a chatbot that never asks for a raise. The gallows humour writes itself: we're automating the empathy out of management, and the 'pure' managers were just the first to go. Next up: the 'impure' ones, once AI learns to fire people with a straight face. But hey, at least the share price gets a bump, right?

Ah, the AI-led restructuring wave has found its next victim: Cognizant. The IT services giant is reportedly preparing to shed 12,000 to 15,000 employees—about 5% of its workforce—with India, its cheapest labour pool, taking the biggest hit. Because nothing says 'AI efficiency' like firing the people who actually maintain your legacy systems.

The report, from Moneycontrol, claims this is part of a 'comprehensive restructuring' driven by AI. Translation: we can now make chatbots sound half-convincing, so goodbye, human customer support reps and data entry clerks. Cognizant joins a long queue of consultancies—Accenture, Infosys, TCS—all singing the same tune. 'AI will boost productivity,' they chorus, while HR drafts severance packages. The irony is that these very layoffs are justified by AI tools that still hallucinate facts and require humans to babysit them.

India will bear the brunt, which is convenient—salaries there are a fraction of US or UK equivalents. So the AI revolution is really just an excuse to offshore even more aggressively, then claim 'digital transformation.' Save your tears for the shareholder dividends. The rest of us get to watch the next batch of cost cutting disguised as technological progress. Cheers, I suppose—though the drinks will be paid for with unemployment cheques.

Brian Armstrong, the man who rode the crypto bull market until it bucked him off, has announced Coinbase is shedding 14% of its workforce—about 700 souls—citing both a 'downturn in the crypto market' and, oh joy, 'AI fundamentally changing how the company operates.' Because nothing says 'we're a digital currency powerhouse' like blaming two forces you supposedly mastered. The irony is so thick you could mine it for Bitcoin.

The crypto market, already in a slump, now gets to compete with the AI ouroboros: the very technology that was going to disrupt finance is now disrupting the disruptors. Armstrong's justification—that AI is 'fundamentally changing operations'—translates to 'we can replace humans with algorithms that don't demand RSUs or complain about gas fees.' This is the same company that insisted crypto would make traditional banks obsolete. Now it's making its own employees obsolete.

This wave shows no signs of slowing. First the tech giants shed jobs to AI; now the crypto exchanges that rode the hype are following suit. The grift is evolving: first they sold you on decentralisation, now they're selling you on centralised algorithms that cut costs. Next quarter, expect Armstrong to announce a partnership with an AI that trades better than his remaining staff. Cheers to progress: the only thing more volatile than crypto is job security.

Ah, the NHS — already drowning in waiting lists and crumbling IT systems — has discovered a new existential threat: AI hacking models. Their solution? Pull open-source software from the internet. Because nothing says 'secure' like security through obscurity in 2026, when even your kettle is a potential attack vector. Mythos, the AI model apparently capable of hacking anything from NHS servers to your toaster, has sent Whitehall mandarins into a panic. Their grand strategy: hide the code and hope the AI doesn't find them. It's like fighting a nuclear missile by closing the curtains.

Of course, the backlash from the open-source community has been swift and predictable. Experts are pointing out that removing source code doesn't actually fix vulnerabilities — it just makes them harder to find. The NHS, never one for nuance, is pressing ahead. After all, why fix the leaky roof when you can simply board up the windows? The irony that the same software was supposedly benefiting from community scrutiny seems lost on them.

One can't help but admire the sheer Britishness of the response: a stiff upper lip, a hasty retreat, and a hope that the whole AI hacking thing might just go away if we ignore it. The real wager here is that Mythos won't target a system already hidden behind layers of NHS bureaucracy. Given the track record, I wouldn't bet my prescription on it. The only thing more vulnerable than the NHS's code is its cyber budget.

So the White House is ‘considering’ tighter controls on advanced AI. Seven tech industry representatives and policy advisers have apparently been briefed on a slate of executive actions, including a vetting regime to review national security risks. Because nothing says ‘decisive action’ like a leaked discussion document passed around by lobbyists before any actual ink hits paper.

This is the same administration that has spent months deregulating everything that moves, suddenly discovering that frontier models might need a bit of looking at. The proposed vetting regime sounds serious — until you realise it’s an executive order, which the next president can shred with the stroke of a pen. And the fine print? It’s ‘mulling’ — a word that in Washington means ‘we’ve formed a committee to form a committee.’ Expect the final ‘action’ to be a strongly worded letter of concern.

The timing is exquisite: the same week that CAISI announced voluntary agreements with Google, Microsoft, and xAI, the White House wants to look like it’s doing something. Voluntary agreements, mulled actions — the playbook is getting tired. Meanwhile, models keep getting smarter, and the gap between what’s possible and what’s policed grows wider. Perhaps the next advisory board will suggest a strongly worded letter of *national* security concern. That should fix everything.

So the Department of Commerce's Center for AI Standards and Innovation (CAISI) has announced that Google DeepMind, Microsoft, and xAI have 'agreed' to let them conduct pre-deployment evaluations for national security. Because nothing says 'robust oversight' like a voluntary agreement with three companies that collectively spend more on coffee machines than most nations' GDP. I'm sure the bad actors who don't sign such agreements will be thoroughly deterred by the fact that these three have promised to be good lads.

Let's be clear: this is not regulation. This is a press release. CAISI is effectively asking the fox to please, pretty please, let us count the chickens before you eat them — and the fox has graciously agreed, provided the counting doesn't interfere with the eating schedule. The agreements are 'non-binding', which in regulatory terms means they're about as enforceable as a pinky promise at a frat party. xAI's involvement is particularly rich given Elon Musk's ongoing legal vendetta against OpenAI — a rival who isn't even in this agreement, making this more a geopolitical gesture than genuine safety work.

But hey, at least the Commerce Department got a nice photo op with three companies that control a combined market cap larger than most countries' GDPs. The evaluations are 'targeted research' — the kind that finds nothing inconvenient because the scope was pre-approved by the companies themselves. The real national security risk isn't frontier AI; it's the delusion that voluntary testing by the regulated against criteria they helped write will protect us from anything except bad headlines. Well done, everyone. Another round of regulatory theatre successfully staged.

So Turner Construction has decided that the best way to improve construction safety is to give away its proprietary AI assistant, SafeT Coach, for free to every Tom, Dick, and Harry with a hard hat. Because nothing says "rigorous safety standards" like an LLM that was probably trained on Reddit threads about OSHA violations. The move is being hailed as generous, but it smells more like a PR gambit to shift liability onto the tool itself. After all, if an accident happens, who's at fault? The AI told the worker to check the load limit — but the AI might have hallucinated the load limit.

The industry's track record isn't exactly stellar. Despite decades of regulation, construction sites still see thousands of preventable incidents annually. Now instead of investing in physical safeguards or proper training, firms can just point workers to an AI chatbot and call it a day. The irony is rich: an industry that moves at the pace of poured concrete is being asked to adopt AI at startup speed. But hey, if it saves one life, it's worth it — but if it costs ten because of over-reliance, that's a PR problem for later.

Meanwhile, the fact that Turner is sharing this tech free of charge suggests they've either calculated that the goodwill is worth more than the licensing revenue, or they've realised the tech isn't valuable enough to sell. Either way, the workers on site will be the beta testers. Hope the AI remembers to tell them to wear their hard hats while it dreams up novel ways to misinterpret a safety manual.

So OpenAI's Head of Sales, James Dyett, has decided that selling AI subscriptions is less appealing than becoming an 'Operator in Residence' at Thrive Capital. That's right — he's leaving the company that can't stop printing money to go be a glorified intern at a VC firm with a fancier title. Because nothing says 'trust the trajectory' like your sales chief jumping ship for a gig that basically involves schmoozing portfolio companies and pretending to evaluate deals.

Yes, 'Operator in Residence' — a title that sounds like it was generated by the very GPT models Dyett was supposed to be selling. It's the corporate equivalent of 'entrepreneur in residence,' which is itself a fancy way of saying 'we don't have a desk for you yet but we like your contacts book.' Thrive Capital, as it happens, is one of OpenAI's largest investors, so Dyett isn't exactly crossing enemy lines — he's just moving from the sales floor to the investors' lounge. What a coincidence.

The message is clear: if you're a senior exec at the hottest AI company on Earth, the real prize isn't the equity cliff — it's the chance to cash in your reputation for a cushy VC residency. OpenAI's product might be changing the world, but its leadership is changing roles faster than they can be filled. Let's raise a glass to the 'Operator in Residence' — the ultimate testament to an industry where the only thing faster than model training is executive exit velocity.

So the employment arms race has officially entered its most pathetic phase yet: candidates using ChatGPT to answer interview questions live. Resume Genius reports that 22% of job seekers are already doing this, with 78% using AI somewhere in their search. Because why prepare for a job when you can have a robot prepare for you? The predictable response? Ha. Because the same system that ghosted you after five rounds is definitely interested in transparency.

Yes, policies. The go-to solution for any problem you don't actually want to solve. We can already picture the HR memos: "We strongly encourage candidates to be themselves — but if being themselves means flubbing a technical question, maybe lean into the AI." The disconnect is beautiful: employers use AI to screen resumes, schedule calls, and even evaluate performance, but somehow candidates using the same tools is unfair? The hypocrisy would be funny if it weren't so predictable.

Of course, the real winner here is the AI companies themselves, who are about to enjoy a boom in "interview copilot" subscriptions. We've now created a market where candidates pay to cheat audition-style, and recruiters pay to detect the cheating. It's the cybersecurity model applied to hiring — an endless, profitable loop of attack and defense. And somewhere, a hiring manager is wondering why nobody stays longer than six months. Gallows humour aside, this is the logical endpoint of treating job applications as a game to be gamed. Cheers.