UK government replaces Palantir software with internally-built refugee system

Unfortunately, it's hard to escape the feeling that friends in high places, some lobbying and some er... reciprocal back scratching might have been instrumental.

See also senior staff at NHS England (or Digitial? can't remember) handing massive NHS compute contracts to AWS, and then leaving the civil service to become... an AWS employee.

The company bid for this contract, and lost to Palantir. I still can't believe that a company trying to do this in exactly the right way lost to a US intelligence company.

1. they aim to deliver product company margins with a consulting-heavy model.

2. your software purchase funds a cadre of "free" FDEs and deployment strategists who customize your install, build a bunch of data pipes/transforms, and talk to people to figure out what all the data means.

This could be a good deal if your tech org is not very competent at integrating your data, or if you have a sudden, short-term need. In the longer term, it's probably cheaper and more effective to develop a competent tech team, modernize the source data systems, and roll your own integration -- but that also requires leaders with long-term vision who are resistant to external hype and pressure.

This is the kind of thing GDS and other Civil Service departments build all the time, its a completely standard kind of challenge that a small team of Devs (+ supporting staff) from a departments DDAT department does day in and day out.

The output will be open source by default and use existing standards.

For context, the Homes for Ukraine refugee scheme cost 2-3 billion as of 2023. I can't seem to find an updated cost. This cost (from the article) was Palantir working for free for the first 6 months (could they have beat that, time wise?), then awarded 4.5m and 5.5m for two more 12 month terms, and now they're transitioning to something home-grown instead.

> The MHCLG [ Ministry of Housing, Communities and Local Government] said it initially needed a system which could be ready within days but, in seeking a "steadier service", later created an updated platform to meet the programme's longer-term needs and bring down costs.

I basically agree with the MHCLG's reasoning here. It's always worth at least experimenting to see if you can roll your own.

Free software from Palantir is not free. Peter Thiel's co is all about monopolies.

Oh, and don't forget to opt your data out of Palantir: https://your-data-matters.service.nhs.uk/

Tjheir "ELITE" guide says that during "special operations" normal safeguards may need to be turned off.

Palantir's Maven Smart System ha grown into a Pentagon program of record with 20,000+ active users. "Human in the loop" may become "human rubber stamp" when the number and speed of AI recommendations exceed real human review capacity.

A Palantir-backed program reportedly operated secretly from city council members, defense attorneys, and the public.

Vendor lock-in issue: once a system becomes embedded in agency workflows, switching vendors becomes politically and operationally hard and they are trying their best to achieve this. The Army's $10 billion enterprise agreement consolidating many contracts into one Palantir platform is the cleanest example of institutional dependence.

--- tldr;

The accountability chain is broken: when harm happens, the agency blames the tool, the vendor blames the customer, the operator blames policy, and the model blames the data.

---

Also, I won't share the full report link since whenever I share something like that here, I get banned/flagged for a day.

https://www.statista.com/statistics/1388245/uk-sick-leave-fi...

Now, I get it. The UK is not exactly providing good jobs for a lot of people, so of course we're seeing this. But, getting a mental health diagnosis, be long-term-sick and avoiding unemployment while both getting paid by an employer and getting unfair advantages out of the public health system ...

In the UK, long term sick leave with a mental health diagnosis is a way to be unemployed (ie. not working) but not have the disadvantages of that. There's the money difference: 530 pounds per month for sick leave, 338 or 425 pounds per month for unemployment. On long-term sick leave you get all advantages job seekers get (ie. "Universal Credit"), PLUS others (support for not being able to work, ESA, and support for extra living expenses due to long-term sickness, PIP). So if you don't want to work, long-term sick leave has many advantages (you can even put that you're working on your CV), plus it's a big cost to employers. You don't have to look for work in long-term sick leave. In fact, nothing is expected at all (other than medical evidence).

2.8 million people are long-term ill, not participating in the workforce, not being economically productive at all, at least half due to being diagnosed with mental illness. The issue with this is that this is happening with full support of government employees, and even the courts cooperate to a lesser extent.

What the government is trying to do, in other words, is trying to kick people off, uh, let's say "mental disability", force them to work. And they need to do this without relying on government workers, because they often side with the people on sick leave.

Hence, Palantir. Being hated is a feature here, not a bug.

I would never trust an openly MAGA company.

(FTFY)

In my own experience consulting with factory owners—advising them on hardware choices like Lumens versus Mitsubishi—I see the physical reality. The idea of absorbing a client's database into an ontology and hooking it up to an LLM sounds great in theory. However, considering the extreme fragmentation of equipment standards and data representations across different sites, I seriously question if this is a sustainable business model.

Sure, initially it’s just dispatch programming. But how can they possibly absorb all these disparate, chaotic field environments into a single platform asset? Even within a single factory, different assembly lines use entirely different equipment, often from completely different manufacturers.

The idea of interpreting every piece of equipment's specific protocol, reverse-engineering the DB schemas, standardizing the terminology, and modeling the entire approval flow seems practically impossible. Is this actually achievable? Take PLCs, for example: even if they share a standard communication protocol, the ladder logic itself is completely incompatible across different brands

Thinking about it in reverse, Palantir might have absolutely no intention of solving this fragmentation problem themselves. Their survival strategy might be to dictate the core tech stack of the end-point B2C clients, creating a structure that essentially incentivizes specific B2B vendors to fall in line. Ultimately, what makes Palantir so dangerous is the high likelihood that they will simply shift the massive cost of standardization onto those B2B subcontractors