The New Dark Age

Saturday, 10 July 2026 — medConfidential

NHS 2028 – The Trajectory of Current Decisions (from medConfidential)

The Health Bill/Act of 2026 is expected to come into force for 2027, creating an all-powerful Department of Health in England which will have powers of Direction and delivery unmatched since the creation of the NHS. Delivery will be steered by DH/E AIs.

The NHS is the largest part of the State that cares about people. The managed decline of the NHS and GP is brought to you by Palantir and the consultants who know their own private healthcare will shield them from the consequences of their actions to implement choices that Mr Mandelson encouraged his political protege to make. Austerity and cuts have shrunk all the other parts of the safety net, those that would previously have seen someone was starting to struggle but before they hit rock bottom of clinical intervention. Now the assessment of rock bottoms will be what a patient is willing to type into an AI text box (characters limited) to see if Palantir’s machine thinks are worthy of care; everything the patient types will be summarised and stored for future retrieval and may be used against them when the algorithms decide to dredge it up again years in the future.

There’s no evaluation, no critical event audits, and the App AI will be rushed out without detail to inform the coroner reports. So those things will have to come in reverse order. We’ll be here.

In the interim, GP will be liable and responsible for data access via the SPR, with no ability to do anything about it and no way to meet their responsibilities when DH/E facilitates abuses by creeps and ghouls.  As with online triage, GP will be punished for what they are not permitted to do, and used as an excuse to dismantle the entire system.

In many ways, the Department of Health in England will return to the worldview it had when it was DHSS and mostly ran benefits with the NHS attached as an afterthought. It’ll have a network of public-facing hospitals, (Community) Diagnostic Centres, and “neighbourhood health centres” with the footprint roughly equivalent to that of a JobCentrePlus. They’ll treat “visitors” similarly.

Announcements so far

• The Single Patient Record running on the Federated Data Platform from Palantir. It’s notable that this first key point is the one that the Department of Health in England are most vague on, to the point that they may have misled Parliament. Not that Palantir care about things like that. SPR is the key piece underlying everything else – Palantir may be cartoon baddies, but they can be replaced quite easily and the below agenda continues unchanged.
• National Online Hospital – expected to take all conditions that don’t always need physical context. It won’t handle your broken leg (A&E), and you’ll still need to go somewhere for an x-ray (Community Diagnostic Centre), but any conversation that could be a call will become a call.
• Banning walk-in to A&E – if you want in person care, the AI/algorithm in the app has to have said yes. If you try to walk into A&E while bleeding, the iPad at the door will decide if you can continue. It’ll know everything in your SPR, including what it decided you didn’t need that morning, and it won’t let you forget it.
• GP goes app first – it is the old “GP at Hand” on steroids, but as it is a political imperative that it must not be allowed to fail, there will be no alternative.
  • DH/E will already control the phone lines of GP practices, and it already controls the registration process, and online triage is micromanaged.
  • Patients who don’t or can’t use the app will be delayed by the algorithm behind those who do.
• “List cleaning” will take signals from the NHS app, so if patients use the app less than average, the practice gets punished more than average.
• More than half of GPs will close – When most interactions with GP are via the NHS App, and patients are directed away from visiting the practice, ICBs will see GP premises as an expense which Managing Public Money says should go elsewhere from 2028. Closures of GPs will be met by more MP consternation than when bank branches and post offices close, but their arguments change nothing – DH/E expects ICBs to take the same approach when they defund GP practice buildings in favour of spending the money on the neighbourhood health centres and community diagnostic centres owned and managed by hospitals. Patients and the public may complain, but the Streeting/Mandelson vision is that they can howl at the moon (Mandelson never cared about reelection).
• When the GP closes, many local pharmacies will go too.
• (Rural practices where the ICB/Trust is now far away may survive for longer, practice partners will largely go bust (unlimited liability risks their homes and savings). For GP owners, there is a first mover advantage and a very large late mover disadvantage to the new world. Those who escape while GP is still gliding on will get maximal benefit – the business will become de facto worthless over a short time unless independent revenues are available, or the building can be sold for flats.)
• The replacement of practices with the App/AI overlord will become the playbook for the AI bros to move on to the consultants.
• The least affected will be the junior doctors – although there’ll be far fewer places to progress through to, just in time for the covid-inspired cohort to emerge from medical school.
• It is recognised that access to mental health support will depend on being able to clearly admit and articulate need in a way which the AI finds believable. If the AI believes the text it is shown, and takes into account past history, the patient gets waved through at every stage. If not, the patient gets blocked at every stage. Access to care will become based on access to advocacy, with all the perverse incentives towards ever more extreme acts that come with such gatekeeping.

All these press releases will be the citable excuses that DH/E use to say they couldn’t possibly use the Palantir break clause. Holding cancer patients hostage wasn’t working for DH/E and Palantir, so they have switched towards all GP patients via the App AI – which runs in Palantir and no one external will know enough about how it works until after Palantir’s sales team have insisted that only Palantir can possibly do it (others can, the reason DH/E hear that Palantir are the only ones who can do it is because Palantir are the only ones that DH/E invite to those meetings). The only time that an announcement doesn’t include supplier quotes is when DH/E are embarrassed by the supplier.

• The referral checking AIs will also gatekeep the online hospital, as a mechanism to keep waiting lists down. In effect waiting lists will be zero because they’ll be hidden. Patients aren’t waiting if they’re in managed care (and over time the later steps will become less and less resourced because it’s clearly fine without them).
• Populating the SPR repository via direct care channels will avoid the GP Data Opt Out, after which DH/E will continue to argue it doesn’t bind them. SPR’s IG will argue it will inherit Summary Care Record and Shared Care Record infrastructures, but will wipe out the patient objections because the “privacy, transparency and trust” office role is to ensure that none of those things apply to patients. All the medical notes, prescriptions, diagnoses and service accesses will be made available to research via the HDRS, and DH/E will continue to argue that in the name of privacy, transparency and trust that no opt outs will apply when DH/E and HDRS sell patient data – it might harm the profit margins.

What’s next

• App first allows for entirely different models of care, the way there are entirely different models of banking. But DH/E and the app will require political approval for innovation in areas and choices about the prioritisation and privatisation of care. The SPR will be privatisable under a future Government.
• App triage can be good. There are many innovations in health that could be done in the app – testing breathing via an on-device AI stethoscope would be a reassurance for those who don’t want to bother the NHS at 4am in the morning unless they have to. Knowing that lung function is better or worse than yesterday can prevent a 999 call at 4am, giving a fearful pensioner the reassurance to go back to sleep as they’re no worse than previously, or the information they need to wake up their partner, without having to bother anyone unnecessarily. But none of this is in the roadmap because it doesn’t directly help the DH/E budget (it only helps ambulances and A&E).
• The good intention is for the app to become a lifelong companion, but there’s a fine fine line between companion, carer, and overseer. What started for emergency inpatient mental health turned into permanent surveillance everywhere that Oxevision normalised across the NHS. The approach will be repeated by other suppliers for those conditions where the NHS handles the consequences of human behaviour and politicians choose to micromanage it in return for care or medicines.
• Modern devices have many sensors which store data on health (and health proxies). If a person chooses to wear an Apple Watch for one reason, a future app update can require they must share all health data with the NHS app in order to continue to receive care.
  • It was only the DH/E that used an all or nothing approach to the covid app – if the user didn’t give it all the sensors and data that it could use, there was a political decision that it would refuse to work. Everywhere else in the world did what they could with the permissions given, but DH/E demanded everything.

• Clinical tests will become subject to political priorities. For example, at the national level, the PSA test is not value for money, but that doesn’t matter to an individual person (or the charities fundraising off them) who want to prioritise those topics. Patient advocates don’t give up.
  • By the end of this Parliament it will be possible to go to the vet to cure some cancer in your cat (not all cancers, not all cats). The notion of political priorities will allow the app to facilitate a blood test (that works) for cancer at the pharmacy, and pop back a week later for the mRNA jab to treat that cancer. Whether the NHS does that (and whether patients will be required to have their DNA stored forever in Palantir) is an unanswered question, but this is a roadmap that was previously nationally difficult.
  • SH:24 continues to be a national treasure broken up by postcode lotteries – how the new DH/E deals with that will show how everything else goes (and how they deal with it continues to be a postcode lottery).
• DH/E algorithms and decisions will be political – trans care and ADHD as the canaries in the coal mine, written to your SPR forever and the algorithm can just say no. The dalliances of youth will be recorded and marked on your permanent record forever. The precedents of trans care and ADHD prescribing will get progressively wider as the culture war needs new victims. Any pressure group will demand the system bend to their beliefs – bigots and obsessives don’t give up (and can masquerade as patient advocates in some contexts).

• Gatekeeping – GPs have been told the AI will be gatekeeping patient access to GP, and GPs have also seen the “single point of access” as gatekeeping GP referrals to hospitals.
  • The ban on referrals that started with GP in the 26/27 contract will return as GPs close, and then will roll on further into the NHS. Professionals will need to convince the machine that the patient need is greater than the political desire to fiddle waiting lists, and only in cases where the clinician’s commitment to their patient extends higher than the barrier erected to referral will the referrals go through. Such barriers will reappear at every step
  • Eventually the algorithm will decide that you don’t need the ambulance or the trip to A&E, and Palantir’s reimplementation of the US healthcare system will be ready to be sold back to US health insurers. Of course, few other places in the world would accept something from the current NHS incumbent supplier.

• PIP appeals will be harder. If a patient can’t get a GP appointment without the AI approving it, they’ll never get the necessary information and paperwork to challenge a DWP assessment – and DH/E will share the NHS data with DWP as soon as DH/E needs DWP support in a political quid pro quo.

• As the app gets reworked to be a “native” app, it will require suppliers to support at least three different interfaces (they’ll need a web interface for a while, plus iOS native, and Android native). As tech interfaces shift, the NHS App will remain with a legacy that does not allow it to follow; the service will simply be seen as old and broken. But it does make one particular DH/E app team’s life easier for now, and that team has imposed that vision on all patients and suppliers without anyone externally getting a say. This is how DH/E app development goes wrong.
  • It’s unclear how the app can work for a 13 year old to manage her own health if she’s banned from having a smartphone until aged 16.

• Patient notes will document what happened after the patient dies, if any coroner is allowed to look (someone should write some questions for them!).
• Nowhere in the roadmap is app triage functionality to help patients, not institutions. It’s the sort of thing a third party could offer if they wanted to, but DH/E doesn’t play nicely with others and repeatedly destroys innovation. The gorilla has got its statutory powers and can tread on whomever it chooses. It is going to.

What to do about it

You paying for the medConfidential Substack would be a start. Because no one else is coming to save any of you.

All Governments end; although whether institutions can survive such a comprehensive push for abolition is unclear. Those that may survive longest may not be those that should survive longest. The options that will survive inside DH/E must be those palatable to the Palantir and civil service worldviews.

Nowhere was the above debated before it was announced – it was concocted with the preferred supplier and Streeting in opposition when Mandelson was flying high in the Labour party. As with the now disgraced dark Mr, he knew what the public reaction would be, so hid it all until he hoped it’d be too late. How’s that going to go this time?

Now it is time for implementation. The SRO for the FDP and SPR insists to Parliament no decisions have been made about storing the SPR in Palantir. When full facts emerge it’ll be clear he was probably disingenuous at best (his predecessor knew not to talk publicly about closing GP practices like post offices and bank branches; the current NHSE chair did anyway).

If Andy Burnham says that he stands for anything, it’s hope for the future, over the managed decline that Treasury, Mandelson and Streeting have chosen for everyone else.

The above is the current trajectory, it is not set in stone. Political will can change things, and there is change in the air. There are adjacent alternative visions of the future – just because they’re currently entirely ignored and disconnected from political imperatives does not make them incorrect.

If the Mandelson-Streeting-Palantir worldview is fully implemented, perhaps the advice may have to become: “Don’t get sick, and if you do get sick, die quickly.”

medConfidential