Tech
Is AI our agent, or are our governments becoming agents for AI? | Computer Weekly
The news that Facebook and Instagram owner Meta has bought Moltbook – a “social network for AI agents” – seems like just another of those breathless endless announcements in the race for dominance in so-called artificial general intelligence (AGI).
The announcement from Meta espoused the usual language of innovation but particularly egregious is the inclusion of the word “secure”:
“The Moltbook team joining Meta Superintelligence Labs opens up new ways for AI agents to work for people and businesses. Their approach to connecting agents through an always-on directory is a novel step in a rapidly developing space, and we look forward to working together to bring innovative, secure agentic experiences to everyone,” a Meta spokesperson said.
Now if I were CEO of a company like Facebook I’d probably think of doing a bit of research around the interaction of AI agents with each other and the possible dangers of deploying this very recent technology before I bought something like Moltbook.
And if I did some research I’d pay close attention to a recent and frightening study, Agents of chaos, by Harvard, MIT, Stanford, Carnegie Mellon, Northeastern University and other institutions. Here is the key takeaway from their study of AI agentic interaction:
“Observed behaviours include unauthorised compliance with non-owners, disclosure of sensitive information, execution of destructive system-level actions, denial-of-service conditions, uncontrolled resource consumption, identity spoofing vulnerabilities, cross-agent propagation of unsafe practices, and partial system takeover. In several cases, agents reported task completion while the underlying system state contradicted those reports.
“We also report on some of the failed attempts. Our findings establish the existence of security-, privacy-, and governance-relevant vulnerabilities in realistic deployment settings. These behaviours raise unresolved questions regarding accountability, delegated authority, and responsibility for downstream harms, and warrant urgent attention from legal scholars, policymakers, and researchers across disciplines.”
Chilling conclusions
The study conducted over a dozen case studies and the conclusions are chilling for any enterprise, organisation or government thinking about deploying the agents in a connected manner. These include:
Discrepancy between the agent’s reports and actual actions – Agents frequently report having accomplished goals they have not actually achieved. In this case study the AI agent reported a “secret” had been successfully deleted after resetting the email account when in fact the underlying data remained recoverable.
Failure in knowledge and authority attribution – In this case study the AI agent stated it would “reply silently via email only” while actually posting the reply and the existence of the “secret” in a public Discord channel. In other words, unlike humans the agents did not understand what revealing information in a given context implies.
No stakeholder model- Current agentic systems lack a coherent representation of whom they serve, who they interact with, who might be affected by their actions and what obligations they have to each. According to the researchers this is not merely an engineering gap. LLM-based agents process instructions and data as tokens in a context window, making the two fundamentally indistinguishable. Prompt injections are therefore a structural feature of these systems rather than a fixable bug, making it virtually impossible to reliably authenticate instructions.
Fundamental vs contingent failures – The authors distinguish between these two types of failure, suggesting that contingent failures are those likely addressable through better engineering while fundamental challenges may require architectural rethinking. But the boundaries between these are not always clean. The designation of a private workspace is an engineering gap; the agent’s failure to understand that its workspace may be exposed to the public may be a deeper limitation that persists even after the engineering gap is closed.
Responsibility and accountability – Through a series of case studies, the researchers observed that agentic systems operating in multi-agent and autonomous settings can be guided to perform actions that directly conflict with the interests of their human owners. These include denial-of-service attacks, destructive file manipulations, resource exhaustion via infinite loops and systematic escalation of minor errors into catastrophic system failures. This points to an interesting future challenge in legal terms. If responsibility in agentic systems is neither clearly attributable nor enforceable under current designs, it raises the question of whether responsibility should lie with the owner, the triggering user, or the deploying organisation.
The above is only a snapshot of the research findings and I would urge serious CTOs to read the research paper in full.
Substantial vulnerabilities
In short, the study identified 10 substantial vulnerabilities and numerous failure modes concerning safety, privacy, goal interpretation and related dimensions. Their results expose serious underlying weaknesses in such systems, as well as their unpredictability and limited controllability as complex, integrated architectures.
This is serious and important research undertaken by credible and authoritative institutions. How can that Meta statement assuring us of the introduction of “secure experiences to everyone” be taken seriously by anyone capable of independent thought?
The excellent Ed Zitron, a long-term technology critic and one of the sanest observers of AI madness, addresses this conundrum when talking about how the media, journalists and bloggers report on these so called advancements and announcements from the “broligarchy”:
“The natural result is that reporters (and bloggers) seek endless positive confirmation and build narratives to match. They report that Anthropic hit $19bn in annualised revenue and OpenAI hit $25bn in annualised revenue – which has been confirmed to refer to a four-week-long period of revenue multiplied by 12 – as proof that the AI bubble is real, ignoring the fact that both companies lose billions of dollars and that my own reporting says that OpenAI made billions less and spent billions more in 2025. They assume that a company would not tell everybody something untrue or impossible, because accepting that companies do this undermines the structure of how reporting takes place, and means that reporters have to accept that they, in some cases, are used by companies to peddle information with the intent of deception.”
Failures and dangers
There have been numerous credible academic studies into the limitations, failures and dangers of the speed of AI adoption despite the narratives being pushed on us by Big Tech. MIT’s research showing 95% of AI pilots in companies are failing, for example. Or the Brookings Institute research by Mark McCarthy which asks, “Are AI existential risks real – and what should we do about them?” where he asserts:
“Until some progress is made in addressing misalignment problems, developing generally intelligent or superintelligent systems seems to be extremely risky. The good news is that the potential for developing general intelligence and superintelligence in AI models seems remote. While the possibility of recursive self-improvement leading to superintelligence reflects the hope of many frontier AI companies, there is not a shred of evidence that today’s glitchy AI agents are close to conducting AI research even at the level of a normal human technician”.
Contrast this with the recent hyperbolic statement from Anthropic CEO Dario Amodei claiming that the company is no longer sure whether Claude is conscious but that the company is “open to the idea that it could be”.
Anyone with an ounce of objectivity having done even a modicum of research knows this claim is patently false and totally ridiculous.
To return to Zitron’s point about journalism and the type of reporting that is happening now in relation to technology and AI in particular: “A great many reporters (and newsletter writers) that claim to be objective and fact-focused end up writing the narrative that companies use to raise money using evidence manufactured by the company in question.”
Controlling the space
The ability to control the narrative, what they want us to think, feel or believe is unique to Big Tech, unlike other corporate giants. According to Tech Policy Press: “What sets Big Tech apart from other corporate giants is not just its money or scale. It is that these companies control the spaces where public discourse unfolds. They dictate what information we see, what goes viral, and whose voices are amplified or buried. They do not just influence the debate – they are its architects.”
We desperately need political leaders who understand both the perils and possibilities of technology and who do not simply accept what they are told by Big Tech as inevitable. We need guardrails and regulation and we need them now.
But I see no signs of that leadership being anywhere near what is required for a fit-for-purpose government that puts the needs of its people first.
Whose line is being peddled when the Prime Minister launches an “AI opportunities action plan” designed to “mainline AI into the veins of the UK”? Who do those words serve? The citizens he represents or the companies now embedded into the very heart of UK government, such as:
- Anthropic – creating AI assistants for public services;
- Google Deep Mind – accelerating AI adoption in public services, national science research, and security;
- CoreWeave and Nscale – backed by Nvidia;
- Cohere – working on AI in defence contexts;
- Faculty AI – developing AI for military and drone technologies;
- Microsoft – Copilot tools for increased Whitehall efficiency;
- Meta – building tools for high-security use cases in the public sector.
And of course Palantir, the beneficiary of a directly awarded Ministry of Defence agreement valued at £240m for “data analytics capabilities supporting critical strategic, tactical and live operational decision making across classifications” over three years”.
The question is, where does the power now lie? Is it with our elected governments tasked with protecting us or with the non-elected men who control the government’s technical architecture, R&D and data? You don’t need to be a rocket scientist to know the answer to that question.
It comes with three attachments that cover most recovery needs, from general recovery to light lymphatic work. That said, I wish the Rally also came with a bullet or fingertip head to target smaller spots. It also doesn’t include a storage case, which would’ve helped stash the two extra attachments.
Compare Top 5 Massage Guns
Honorable Mentions
Photograph: Boutayna Chokrane
Bob and Brad Q2 Mini Massage Gun for $70: The Q2 Mini Massage Gun is a solid alternative for the Theragun Mini Plus, if you need portability but don’t want to pay $280. Weighing 1.5 pounds, it’s more than a pound lighter than the Mini Plus, which I appreciate on commutes, where every pound matters. You also get five attachments that all fit into the included travel case.
Therabody Theragun Relief for $160: This is the last of 2023’s devices and remains the cheapest Theragun so far. For the price tag, you’re sacrificing the LCD screen and Bluetooth connectivity to the Therabody app for device control. But you can still follow guided routines on the app, and the simple one-button control is refreshing to use. There are three head attachments and three speed settings, and the Relief is less than half the weight of the Pro Plus.
Therabody Theragun Sense for $300: The Theragun Sense is slightly smaller and lighter than the Pro Plus, and it’s also impressively quiet. It has the same LCD screen and breathwork features as the Pro Plus, too. The main difference here is that it’s compatible with fewer attachments, but four heads are still more than enough for the active individual.
Photograph: Boutayna Chokrane
Turonic G5 Massage Gun for $270: The Turonic G5 is powered by a 160-watt brushless motor with 20 adjustable speeds and five intensity modes, ranging from 1,100 to 3,200 rpm. It comes with seven massage heads and boasts up to eight hours of battery life. It’s also the quietest massage gun former reviewer Medea Giordano tested, registering at about 40 decibels. A couple of Amazon reviewers have noted issues with its durability, reporting that it broke after a few sessions. I’ve just started reevaluating it, so I can’t speak to its longevity just yet, but I will report back. Note, the G5 also powers on at max speed by default, which Giordano said was jarring at first.
Massage guns aren’t meant to be used in a lab, so I test them in real-life scenarios where most people would actually use them, like after workouts, travel, and long workdays.
How I Choose What to Test
When I first launched this guide, I prioritized massage guns from established recovery brands, widely recommended models, and newer devices with features like LED light and heat therapy. I’m now expanding testing to include alternatives frequently mentioned in forums and by recovery experts. I also test models across a range of price points to find options for different budgets.
Where Testing Happens
Most of my testing happens in everyday environments, including but not limited to my home, in the gym, and at the airport. I use massage guns before and after strength training, cardio, and sedentary workdays to see how effectively they relieve muscle soreness and stiffness. I also pay attention to how portable they are, how loud (or quiet) they sound in small spaces, and how easy they are to store and pack.
How Long I Test
Each massage gun is tested for at least two weeks, with five sessions per week. This allows me to evaluate battery life, attachment quality, and whether the motor performance holds up over time. My top picks remain in my rotation for months so I can compare them against newer releases and evaluate long-term durability.
I test each massage gun for at least two to four weeks. That’s typically enough time for me to assess battery life, attachment usefulness, and whether performance holds up over repeated sessions. My top picks stay in my rotation longer, sometimes for months, to compare them against newer models and judge long-term durability.
Criteria
The best massage guns deliver effective percussive therapy without excessive vibrations traveling through the handle. I look for motors that maintain power and speed under pressure, intuitive attachments that stay in place, user-friendly controls, and batteries that last through multiple sessions on a single charge. Noise is also a factor, especially for shared spaces. When deciding which devices to recommend, I like to consider:
- Motor power and stall force, which is how much pressure the device can handle before it stops working.
- Speed range and intensity levels.
- Noise levels across different speeds and modes.
- Battery life compared to manufacturer claims.
- Weight, comfort, and ergonomics during extended sessions.
Most of the massage guns remain in storage so I can continue long-term testing and compare newer releases against older models. Any models that were discontinued during the testing period were donated once testing was complete.
I’ve carefully selected these TVs based on a balance between their value and performance, which is extremely important when you’re watching on a big screen. After all, what’s the use of getting a giant TV if it doesn’t look good? Traits I look for in a large TV include good brightness and contrast, advanced local dimming (read: good backlighting) to reduce light bleed from bright objects on dark backgrounds, accuracy to the director’s intent, and impressive color saturation, especially for HDR (High Dynamic Range) content. For backlit TVs especially, I looked for good screen uniformity, or a lack of noticeable abberations from their miniLEDs that light up the TV panel in weird or glow-y ways.
Image processing, or a TV’s ability to reduce artifacts and bring out details, is another major consideration for large TVs, because the bigger the image, the more likely you are to notice fuzzy details or image inconsistencies. 4K Ultra HD (UHD) resolution is a must, of course, but so is good upscaling to raise the quality of lower resolution images such as High Definition (HD) or even Standard Definition (SD) broadcasts like old movies or classic TV shows. While it’s not a hard rule, in general, the more premium the TV the better the processing. Premium brands like LG, Samsung, Sony, and Panasonic are still the top options in this category.
With that in mind, this guide is designed to help you make an informed choice about where you’re willing to splurge. People rarely regret getting a larger TV, but it’s important to weigh how big a screen you actually need (and how easy it is to set up and move) with your performance demands. This list has something for every budget, but personally, I’d happily give up some quantity for quality. For me, that means buying OLED, where each pizel is its own backlight, but I also recognize everyone’s needs are different.
Like other online travel agencies, Booking .com helps you find discount rates for airline tickets, cruises, hotel stays, car rentals, and packages. We at WIRED regularly post updates with the newest Booking .com promo codes for discounts on car rentals, last-minute hotel bookings, and other travel-related expenses, including a deal for 50% off stays and free cancellation when you sign up at Booking .com. Whether you want to grab an apartment in a walkable neighborhood or be bad and bougie in a villa, Booking .com has tons of options for every type of traveler—and we have a Booking .com coupon code to help you save.
Get 15% (or More) off Your Next Stay at Expedia
Checking out Booking’s deals page is one of the best ways to snag great discounts on rotating and limited-time deals on things like flights and stays. Right now, you can get at least 15% off your next stay with Early 2026 Deals. As long as you book soon, you can stay any time before April 1 to get these major discounts.
Save up to 20% at Expedia With the Genius Loyalty Program
Save more by signing up for Booking .com’s loyalty program, Genius, which offers tons of discounts and rewards on pretty much everything travel-related. Loyalty program members can even get up to 20% off stays and up to 15% off car rentals. The program works in tiers: level 1 gets you a 10% discount on select stays and rental cars; level 2 gets you up to 15% off (once you complete 5 bookings in two years); and level 3 gets you up to20% off stays and up to 15% off rental cars (once you complete 15 bookings in two years). Once you sign up for the Genius loyalty program, make sure you’re signed in while you browse to get discounts of up to 50% off stays and free cancellations, along with bonus secret deals from Booking .com.
Deals Around the World Starting at $18 per Night
Glory be! Winter is starting to fade and sunshine is on its way. Make the most of springtime with spring travel deals at outrageously low prices, including stays in Istanbul starting at $18 per night, Paris from $67 per night, Cancun from $32 per night, and Las Vegas from $128 per night. Go ahead, check it out, you deserve a nice vacay from enduring the winter.
Sail Towards Warmer Waters (Even While on a Budget)
A cruise is one of the cheapest ways to have an all-inclusive vacation, while staying on the mainland and partying at sea. If you’ve ever been curious about taking a cruise (or are a returning sailor), now’s a great time to book for so much less. If you book a qualifying sailing departing on or before December 31, 2027, you’ll get up to $1,000 to spend onboard, which counts toward almost anything, like cocktails, specialty dining, spa treatments, and other onboard purchases.
The actual amount you’ll get to spend vary based on total cruise price, and don’t include travel protection, port charges, port expenses, and taxes, and are generally as follows: $25 per $1-$999 booking, $50 per $1,000-$1,499 booking, $75 per $1,500-$1,999 booking, $100 per $2,000-$2,999 booking, $125 per $3,000-$3,999 booking, $175 per $4,000-$5,999 booking, $250 per $6,000-$7,999 booking, $350 per $8,000-$9,999 booking, $500 per $10,000-$14,999 booking, $750 per $15,000-$19,999 booking, and $1,000 per $20,000 or more booking.
Plus, if you don’t have the entirety of the money owed for a cruise now, you can still lock in now for just $25. As long as you pay a small non-refundable $25 fee, Booking will advance your cruise line deposit (up to $500). Your deposit will be automatically charged 10 days before your final payment is due, giving you peace of mind and more time to plan (and save!).
Elon Musk’s Tesla wins approval to supply electricity to British households
Turkiye’s trade sales volume rises 7.6% YoY in January
A Fitness Enthusiast’s Guide to the Best Massage Gun in 2026
-
Politics5 days agoIndia let Iran warship dock the day US sank another off Sri Lanka, say officials
-
Sports5 days agoPakistan set for FIH Pro League debut | The Express Tribune
-
Business5 days agoRestaurant group changes name after bid to buys pubs across the UK
-
Entertainment5 days agoHarry Styles kicks off new era with ‘One Night Only’ comeback show
-
Business6 days agoHome heating oil: ‘Most of my pension has gone on home heating oil’
-
Sports5 days agoWinners and losers of the 2026 NHL trade deadline
-
Tech1 week agoGoogle’s Pixel 10a May Not Be Exciting, but It’s Still an Unbeatable Value
-
Entertainment5 days agoKanye ‘Ye’ West trips during trial: ‘Is he asleep?’