Every Foundry4 insight, newest first
Everything Foundry4 has published, in date order, across eight sections. If you are looking for a subject rather than a date, the section hubs carry an overview of what is changing in each and why it matters to a British organisation.
What is in here
Foundry4 publishes on eight subjects: artificial intelligence and automation, enterprise software, cloud and infrastructure, data and analytics, cybersecurity, digital transformation, emerging technology, and the technology British public services depend on. The split is editorial rather than commercial. It exists so that a reader who arrives for one procurement question is not served a general technology feed.
Three subjects also have their own hubs because they cut across those sections rather than sitting inside one of them. Health technology touches procurement, data and legacy systems at once. Leadership is a question about decisions rather than about tools. And automation spans a decade of rules-based deployment and whatever the current generation of models turns out to be worth, which is why the site keeps our intelligent automation guide as a single long argument rather than a series of updates.
How to read what is listed
Ordering is by publication date, newest first. Every piece carries the date it was published and the date it was last revised, both as fixed values rather than as something the site recalculates when it rebuilds, so a date that has not moved genuinely means nothing has changed.
Figures carry an inline link to the body that produced them: the audit office, the department, the statute, the filing, the study. Where a number cannot be verified against a primary source, the piece says what is known, describes how a reader could establish the rest for their own organisation, and names who would have to publish it. That is a slower way to write and it produces fewer confident sentences. It is also the only version of this work that survives somebody checking it.
Articles are published under the masthead or under a named desk rather than under a byline. There are no author profiles on this site because there are no individual bylines to profile, and inventing one would be a fabrication of the same class as an invented statistic. The full sourcing and corrections policy sits in the editorial guidelines.
Why some addresses look older than the article
A number of pieces sit at addresses that predate this publication, including two that carry a year in the path. The domain was used by a different organisation before it lapsed, and other sites still link to those addresses. Rather than break the links, Foundry4 keeps the addresses live and publishes a new article on the subject the address names. Nothing from the previous owner is reproduced, restored or continued, and the two dated pieces are written as retrospectives that assess what those forecasts got right rather than as forecasts republished under a current date.
Latest in Insights
London has the data. Citizens rarely see the benefit
The charging network runs on data, not batteries
Who actually moved back, and what it saved them
Most AI projects fail one layer below the model
You cannot fine-tune your way out of bad records
There is no first time. There is only the rehearsal
Hybrid stopped being the thing you settled for
Counting the open data portals still being updated
Open data began well. Then it stopped
The constraint on British compute is electricity
Scoring eighteen predictions from 2018
Product management, in about three minutes
The five RPA applications that survived contact with reality
The five business uses of virtual reality that stuck
Six things a modern government has to get right
The companies building the instruments of the deep sea
Marking nine 2019 technology predictions, seven years on
What the Online Safety Act changed, and what it did not
What enterprise AI agents actually shipped in 2026
Phishing got fluent. Training did not
Most agile transformations stop at the stand-up
The ledger holds. Everything around it does not
Why departments keep merging digital into technology
The job that replaced the transformation CTO
Digitising, digitalising, and the difference that decides a budget
Digital is not what holds the public sector back
Governing enterprise AI without a UK AI Act
The buying process nobody puts on the website
The solution is the last thing you should pick
What the Procurement Act changed for suppliers
The expectations Covid set, and who still meets them
Nobody embraces change. They agree to try it
Social value is scored. It is rarely measured
The public sector's data problem is not about data
Speedboat or oil tanker, and why the answer changed
The quiet crisis in local authority software
NIS2 reaches British companies that never opted in
Following the NHS technology fund to the ward
Platform teams fail for organisational reasons
Britain's quantum position, measured against the money
What a decade of RPA taught us about agents
Per-seat pricing is dying more slowly than advertised
Britain's safety tech sector grew up in public
Councils are being breached through their vendors
The misinformation playbook written in 2020
Who actually runs the world's bike share fleets
What delivery managers actually spend their week on
Subscriptions stopped being a growth story
Pilots clear. Rollouts do not
What remote veterinary triage learned the hard way
What automating a payment run actually involves
The second year is where AI automation gets expensive
Norway's lead was bought, not invented
The build or buy question has a boring answer
The parts of zero trust that survived the budget