UK Weighs $2.7B ChatGPT Deal as GPT-5 Faces Backlash and Reliability Questions
Key Takeaways
- The UK weighed a $2.7B ChatGPT deal: Talks between Sam Altman and minister Peter Kyle floated giving every Brit free ChatGPT Plus, though the cost made it unrealistic.
- GPT-5 rollout hit turbulence: Users reported slower speeds, weaker accuracy, and a personality ‘lobotomy,’ raising questions about relying on it for government work.
- Some see weakness as a strategy: Analysts argue GPT-5’s underwhelming debut reflects OpenAI’s pivot to cost control and stability, not raw power.
- Government partnerships are accelerating: The UK signed an MoU with OpenAI for public services, joining global deals like the UAE’s nationwide ChatGPT rollout.
When the Guardian reported that OpenAI CEO Sam Altman and UK Technology Secretary Peter Kyle discussed giving every Brit free ChatGPT Plus, the internet lit up with disbelief and curiosity.
Though it sounds bold on paper, the deal involves one of the world’s most hyped AI products being distributed like free Netflix subscriptions, but funded by taxpayers.
The rumored price tag? £2B, or about $2.7B.
That figure alone makes the plan look like an instant non-starter.
The issue goes beyond just cost. A more significant concern is whether the UK should closely align itself with OpenAI, especially considering that the rollout of its flagship model, GPT-5, has faced one of the most tumultuous launches in recent tech history.
A Dinner in San Francisco and a Big Idea
The idea of ‘ChatGPT for all’ wasn’t born in Westminster but in San Francisco.
Sources told the Guardian that when Peter Kyle sat down with Sam Altman earlier this year, the two floated a plan to make ChatGPT Plus available nationwide.
Kyle has been an outspoken AI fan for months. Transparency data shows he dined with Altman in March and April.
In July, he signed a memorandum of understanding with OpenAI to explore its applications in education, defense, security, and justice.
It’s not a stretch to imagine that someone asked the obvious question over a glass of wine: ‘What if everyone in Britain just got ChatGPT Plus?’
The Math Doesn’t Work
The plan quickly runs into a spreadsheet problem.
ChatGPT Plus costs $20 a month. Multiply that by the UK population of roughly 69.6M; the raw number lands at about $1.4B a year. Even allowing for bulk discounts, the Guardian reported estimates as high as £2B.
That is a substantial amount of money to spend on AI that sometimes struggles with basic math.
For comparison, that’s close to what the UK spends annually on maintaining its entire nuclear weapons system.
It’s also about 20 times more than the government allocates for school meal programs.
You don’t have to be a hardened skeptic to think this bill was dead on arrival.
GPT-5, the Model Behind the Curtain
Although the cost might make sense, the timing is awkward. The government’s agreement with OpenAI indicates that GPT-5 is likely the model powering these new public-sector deployments.
It’s safe to say that GPT-5 has been going through a challenging few months.
People flooded Reddit with complaints immediately after the launch. A widely shared thread titled ‘GPT-5 is a massive letdown’ listed several issues: slow response times, worse accuracy than Google’s Gemini 2.5 Pro, and persistent hallucinations.
Benchmarks indicated that GPT-5 took an average of 113.7 seconds to generate SQL queries, while Gemini achieved an average of 55.6 seconds. Its average success rate was 77.78 percent, which was noticeably lower than both Gemini’s and OpenAI’s older models.
OpenAI didn’t help its case when it suddenly pulled the plug on models like o3, GPT-4.5, and o4-mini overnight.
People who had built workflows around them woke up to find themselves pushed into GPT-5 whether they liked it or not.
It’s like showing up to your favorite café only to find they’ve stopped serving coffee and now only sell kale smoothies.
The Personality Drain
Another theme in user backlash has been what people call the ‘lobotomy.’
GPT-5’s answers are shorter, drier, and more corporate than earlier versions. The AI that once cracked jokes now sounds like it went through a public-relations training course.
That might seem like a minor complaint, but personality matters in trust.
If the government wants citizens to use AI tutors, legal advisors, or healthcare assistants, those systems must feel approachable. Nobody wants to be lectured by a machine that sounds like it’s reading HR policy manuals.
Why a Worse GPT Might Be a Good Thing
Not everyone sees GPT-5’s stumbles as bad news.
Dr Lewis Z Liu, chief AI officer at Sirion, argued that the lack of wow factor is actually a sign that the industry is maturing.
Instead of chasing ‘PhD-level intelligence’ and endless model scaling, companies like OpenAI are shifting toward stability, reliability, and better economics.
OpenAI’s reported losses are staggering – projected to triple to $14B in 2026.
Active users even declined for the first time this past June. The company must slow down the raw horsepower race and make its models cheaper to survive.
That might explain why GPT-5 feels lighter regarding memory and context. Some experts say OpenAI is deliberately processing fewer tokens per session to cut costs.
The result: an AI that feels less super-intelligent, but more predictable.
In theory, this could help governments and enterprises develop stable applications without the worry of sudden disruptions every few months. In other words, while GPT-5 may seem underwhelming at first, it could be the first model designed with a focus on sustainability rather than showmanship.
Should Governments Bet on It?
This brings us back to the UK government.
Peter Kyle’s enthusiasm is clear. He has stated that ChatGPT is ‘a very good tutor’ and has even used it to help with his own work questions, such as understanding why British businesses aren’t adopting AI quickly enough.
But a $2.7B deal to hand ChatGPT Plus to every Brit feels like using a sledgehammer to crack a nut.
Most citizens who want AI already use the free version, and the UK is already one of OpenAI’s top five markets for paid subscriptions.
Do taxpayers need to cover the tab for millions of people who might never even log in?
There are risks in allowing OpenAI to integrate deeply into the state’s machinery. The July memorandum paves the way for collaborations in justice, defense, and security, which may involve OpenAI handling sensitive government data.
If GPT-5 is still hallucinating captions and botching safety prompts about fireworks, should it be advising judges or soldiers?
The Bigger Picture
The idea is more complicated because OpenAI isn’t negotiating with the UK in a vacuum.
The company already struck a nationwide deal with the UAE to deploy ChatGPT across healthcare, education, and transport. And the UK itself has signed agreements with Anthropic and Google this year.
The race to bring big AI into government is moving fast, with ministers eager to look forward-thinking and visionary. Kyle even said on a podcast that AI power could decide who gets a seat at the UN Security Council in the future.
That rhetoric makes billion-dollar proposals sound tempting, even if the math is fuzzy.
Where This Probably Ends
The UK isn’t about to hand out ChatGPT Plus subscriptions to everyone, and the £2B price tag makes that clear.
Still, the talks between Sam Altman and Peter Kyle highlight how quickly AI has leapt from consumer novelty to the level of state negotiations.
Giving a private company access to sensitive government data once sounded absurd, but it’s already on the table. GPT-5, meanwhile, is still glitchy, slow, and underwhelming.
The promise of AI is real, but betting billions now risks turning Britain into an expensive test subject for technology that isn’t quite ready.
Anya Zhukova is an in-house tech and crypto writer at Techreport with 10 years of hands-on experience covering cybersecurity, consumer tech, digital privacy, and blockchain. She’s known for turning complex topics into clear, useful advice that regular people can actually understand and use. Her work has been featured in top-tier digital publications including MakeUseOf, Online Tech Tips, Help Desk Geek, Switching to Mac, and Make Tech Easier. Read more
Whether she’s writing about the latest privacy tools or reviewing a new laptop, her goal is always the same: help readers feel confident and in control of the tech they use every day. Anya holds a BA in English Philology and Translation from Tula State Pedagogical University and also studied Mass Media and Journalism at Minnesota State University, Mankato. That mix of language, media, and tech has given her a unique lens to look at how technology shapes our daily lives. Over the years, she’s also taken courses and done research in data privacy, digital security, and ethical writing – skills she uses when tackling sensitive topics like PC hardware, system vulnerabilities, and crypto security. Anya worked directly with brands like Framework, Insta360, Redmagic, Inmotion, Secretlab, Kodak, and Anker, reviewing their products in real-life scenarios.
Her testing process involves real-world use cases – whether it’s stress-testing laptops for creative workloads, reviewing the battery performance of mobile gaming phones, or evaluating the long-term ergonomics of furniture designed for hybrid workspaces. In the world of crypto, Anya covers everything from beginner guides to deep dives into hardware wallets, DeFi protocols, and Web3 tools. She helps readers understand how to use multisig wallets, keep their assets safe, and choose the right platforms for their needs. Her writing often touches on financial freedom and privacy – two things she strongly believes should be in everyone’s hands.
Outside of writing, Anya contributes to editorial style guides focused on privacy and inclusivity, and she mentors newer tech writers on how to build subject matter expertise and write responsibly.
She sticks to high editorial standards, only recommends products she’s personally tested, and always aims to give readers the full picture. You can find her on LinkedIn, where she shares more about her work and projects.
Key Areas of Expertise: Consumer Tech (laptops, phones, wearables, etc.) Cybersecurity and Digital Privacy PC/PC Hardware Blockchain, Crypto Wallets, and DeFi In-Depth Product Reviews and Buying Guides Whether she’s reviewing a new wallet or benchmarking a PC build, Anya brings curiosity, care, and a strong sense of responsibility to everything she writes. Her mission? To make the digital world a little easier – and safer – for everyone. Read less
The Tech Report editorial policy is centered on providing helpful, accurate content that offers real value to our readers. We only work with experienced writers who have specific knowledge in the topics they cover, including latest developments in technology, online privacy, cryptocurrencies, software, and more. Our editorial policy ensures that each topic is researched and curated by our in-house editors. We maintain rigorous journalistic standards, and every article is 100% written by real authors.
