Government-Funded Data & AI Training UK

Part of our topic guide on Government-Funded Data & AI Training.

By James Cotton, Founder, iO-Sphere

Most funding explainers list the pots and the rules, then stop exactly where your real decision begins: whether any of it will change what your people can actually do at work. This page runs the decision the other way round. First the map of what the government will pay for, because the phrase "government-funded" hides four different schemes. Then the decision itself, in the order that keeps it honest: name the outcome, choose the route that produces it, and wire the funding to that choice, not the other way around.

A funded course that doesn't change what someone can do is the most expensive training there is, because it spends the dearest resource you have: your team's working time.

The four things "government-funded" can mean

In England, government money reaches data and AI training through four distinct streams. Each has its own funder, its own eligibility test, and its own product shape, and mixing them up is the most common error in funding conversations.

Stream	What it pays for	Who it serves	The test that decides your cost
Apprenticeships (the levy system)	Training and independent assessment on an approved standard, delivered over months alongside the job	New hires and existing employees, any age	Pay bill: over £3m you draw on your levy pot; under £3m you co-invest or receive a transfer
Apprenticeship units (from 28 April 2026)	Short standalone training units, 30 to 140 delivery hours over 1 to 16 weeks, no regulated qualification	Employed learners aged 19 and over	Fully funded for non-levy employers up to the unit's price cap; a levy payer whose fund has run empty pays 5%
Skills Bootcamps	Free or subsidised intensive courses, run by providers in funded cohort windows	Unemployed and self-referred individuals free; employers co-fund for existing staff	Headcount: under 250 employees you pay 10% of the course cost, 250 and over you pay 30%
Adult Skills Fund (formerly the Adult Education Budget)	Funded adult courses, including a free Level 3 for eligible adults through Free Courses for Jobs	Individuals who meet the eligibility rules	Individual eligibility, delivered by providers funded by DWP/DfE or contracted by combined authorities

Two traps sit inside that table. The levy's test is your pay bill, never your headcount: "non-levy" means annual pay bill under £3m, whatever the size of your team. The bootcamp test is the reverse: it is headcount, and the 250-employee line is the one that matters. Carry one test onto the other scheme and you'll misprice your own training.

One more scoping fact worth knowing: the levy is collected UK-wide, but the account you spend from only covers apprentices in England; Scotland, Wales and Northern Ireland run their own schemes. Everything on this page describes the English system, where apprenticeship funding rules sit with the Department for Work and Pensions (since 16 September 2025) and standards are maintained by Skills England.

What is not government-funded: commercial short courses, conferences, tool licences, and bespoke team training. Those are paid products, including ours, and they should be judged and priced as such, never described in levy terms.

Funding rules on this page are stated against the 2026-27 apprenticeship funding rules and were correct as of August 2026. The rules change every funding year and sometimes mid-year, so check the latest DWP funding rules before you commit.

Decide the outcome before you touch the funding

Name the capability you're buying and how you'll see it in the work: a report that didn't exist, a manual process automated, a decision made faster with better evidence. Write that down before you look at a single funding rule. If you can't pass this step, stop here; a funded programme without a named outcome wastes the funding and, worse, the learner's time. A short paid course is the right answer to a vague need.

The pressure to skip this step is real, and it's about to grow. From 1 August 2026, unused levy funds expire after 12 months rather than 24 for new contributions, and "use it or lose it" thinking buys training that sits on a shelf: chosen because budget existed, not because an outcome was named. The expiry clock is a reason to plan earlier, not a reason to buy faster.

The demand side is not in doubt. Around 23% of UK businesses reported using some form of AI by late September 2025, up from 9% when the question was first asked in September 2023 (ONS Business Insights and Conditions Survey, 2 October 2025). Among businesses using AI, the most common workforce response is upskilling rather than replacement: roughly a third train or retrain existing staff, against about one in ten automating roles (same survey). The question facing most employers is not whether to build data and AI capability but how to fund building it in the people they already have.

And think wider than the obvious use. The same funding plumbing pays for graduate schemes run as structured programmes, for retraining people whose roles are at risk, and for productivity programmes in teams that own the data and the manual work: operations, finance, marketing, line management. Same rules, different outcomes, and each decision should still start with the outcome.

Match the route to the outcome

Once the outcome is named, the route usually picks itself.

Durable capability, built over months, evidenced in real work: that is an apprenticeship. An apprenticeship funds structured training inside a published standard, assessed at the end by an independent organisation, for new hires and existing staff alike. It also carries something the unregulated alternatives don't: a regulated framework around what must be taught, how it is evidenced, and who checks it. Where the fit is genuinely there, we point you to the regulated route first, because the regulation is your protection.

A fast, narrow intervention: that is paid training, or possibly a unit. A leadership AI primer, a specific tool rollout, a one-week fluency push for one team: the levy won't fund a short commercial course, and it shouldn't. Our short courses and custom programmes live here, priced as the paid products they are.

Apprenticeship units, live from 28 April 2026, are the funded version of short: 30 to 140 delivery hours, employed learners aged 19 and over, no regulated qualification at the end. We don't deliver units, and our view is that the funding per unit is small and the format constrains depth: a unit suits an employer who needs a fully funded, short, narrow intervention and cannot fund more. Where the goal is genuine capability, a full programme is the route.

An individual without an employer in the picture: that is a Skills Bootcamp. Bootcamps are free to unemployed and self-referred learners and subsidised for employed ones, and they run in funded cohort windows rather than year-round.

We deliver two: a Data Analyst Skills Bootcamp in Oxfordshire, funded through the DWP Skills Bootcamps programme and carrying the NCFE Level 3 Certificate in Data, with a further cohort funded and confirmed for later in 2026 (dates to be announced), and a Data Analytics Bootcamp for Londoners, funded by the Mayor of London's Skills for Londoners programme, which ran in early 2026 and is reapplying for the next funding round. Cohorts run in windows, so check current availability. The individual's view of the same territory, all three pots of public money with their eligibility rules, is our guide to free government-funded data analyst training.

Non-technical teams have their own version of this decision. If the people you're developing sit in operations, marketing, finance or admin rather than a data team, the routing logic is the same but the fit questions differ; we've written that up separately in fully funded AI training for non-technical staff. Data governance capability likewise has its own funded route.

And sometimes the funded answer is no. A funded programme is the wrong choice when you can't name a work outcome, when the learner's manager won't protect the off-the-job hours, or when the need is really a one-week intervention. In those cases a paid short course fits better and usually works better, and the levy will still be there when the conditions are right.

Then wire up the funding

The levy mechanics are simple once separated from the noise. Employers with an annual pay bill over £3m pay 0.5% of it into the levy, offset by a £15,000 allowance, collected monthly through PAYE. That money, ring-fenced for approved training, is what "government-funded" usually means in practice: it is not a grant, and for levy payers it is money already paid.

The Growth & Skills Levy (formerly the Apprenticeship Levy) works the same way, and from April 2026 it also starts funding shorter, more flexible training alongside full apprenticeships; the full detail lives on our Growth & Skills Levy page.

There are three doors to the money, and almost every UK employer fits one of them:

1. Your own levy pot. Pay bill over £3m: training costs come out of funds you've already paid in. For your under-25 hires there's a further sweetener: employers pay 0% National Insurance contributions for apprentices under 25 earning up to £50,270.
2. Co-investment. Pay bill under £3m: you don't pay the levy, and the government funds most or all of the training. Any start you plan now lands in the 2026-27 funding year (starts from 1 August 2026): apprentices aged 16 to 24 at non-levy employers are 100% government funded, so the training costs you £0, and for those aged 25 and over you contribute 5% of the price (DWP apprenticeship funding rules).
3. A levy transfer. A levy-paying employer can transfer up to 50% of its unused funds to any other UK employer, a limit raised from 25% on 22 April 2024 (GOV.UK). Typically it flows to supply-chain partners, SMEs and charities, and it can cover 100% of your training cost: not paying the levy is not the same as not benefiting from it. If you're a smaller employer working out which door is yours, start with our apprenticeships for SMEs guide.

Three other 2026-27 changes are worth planning around. From 1 August 2026 the 10% government top-up on funds entering a levy account is removed, new contributions expire after 12 months rather than 24, and a levy payer whose pot has run dry pays 25% co-investment, for apprentices aged 25 and over only: a 16 to 24 year old is 100% funded even at a levy payer with an empty pot (2026-27 funding rules, v3 amendment of 29 July 2026). Non-levy employers stay at 5% for 25 and over.

Now the other side of the ledger. Funding covers the approved training and the end-point assessment, and nothing else. It will not pay wages, and it will not pay for the working time your people spend learning; travel and kit are also yours. The working time is the cost employers under-count. Against it sits the real value: the government's contribution is capped per standard by a funding band, and on the standards we deliver those bands run from £10,000 to £19,000 per learner. A five-figure programme for £0 or 5% is a remarkable deal; it just isn't a free one, because the time is yours.

The standards you can fund, from Level 3 to Level 7

Levy funding doesn't attach to courses; it attaches to apprenticeship standards: published definitions of an occupation, the skills to evidence, and how they're assessed, maintained by Skills England (which replaced IfATE on 2 June 2025). The standards most relevant to workforce data and AI capability sit at Levels 3 to 5, and those are the levels iO-Sphere delivers. Here is the map, with the government's maximum funding contribution and the published minimum off-the-job training hours for each (source: Skills England):

Programme	Level	Standard	Max funding band	Off-the-job minimum
Data & AI Essentials	3	Data Technician (ST0795)	£13,000	348 hours
Advanced Data & AI	4	Data Analyst (ST0118)	£15,000	370 hours
AI Transformation	4	IS Business Analyst (ST0117)	£18,000	348 hours
Data & AI Strategy	4	Data Protection & Information Governance Practitioner (ST0967)	£10,000	370 hours
Data & AI Governance	4	Data Protection & Information Governance Practitioner (ST0967)	£10,000	370 hours
Data Engineering	5	Data Engineer (ST1386)	£19,000	418 hours

Data & AI Strategy and Data & AI Governance both run on the same standard, ST0967, with a strategy or a governance emphasis: a standard code maps to one occupation, so the identical code is deliberate, not a typo. Every programme in the line was built with AI integrated from the ground up rather than bolted on, because data is the substrate everything in AI runs on. The standards' published typical durations run from 18 to 24 months (Skills England); our delivery runs 15 months of training plus a 3-month end-point assessment.

One current Level 4 standard is deliberately absent from that map: the AI and Automation Practitioner (ST1512), the AI-titled standard, approved for delivery with an £18,000 band. It is genuine, and it is not one we run. Our AI Transformation programme runs on the IS Business Analyst standard (ST0117) because its orientation is applying, evaluating and scaling data and AI within business contexts, where ST1512's focus is automation and pipeline work. If ST1512's exact scope is what a role needs, our Level 4 AI apprenticeship guide weighs the two honestly, including when we would point you at a provider who delivers it.

Above Level 5 the ladder continues: Machine Learning Engineer (ST1398) and the Data Scientist integrated degree (ST0585) sit at Level 6, and the AI Data Specialist (ST0763) is a Level 7, Master's-level standard with a £17,000 band. We don't deliver Levels 6 or 7, and we won't pretend a Level 4 programme is the same thing; if that's genuinely where you're heading, our Level 4 AI Transformation programme builds the applied foundation and a progression route toward it.

There is a higher-stakes reason to check the level before you plan, though: public funding for Level 7 apprenticeships was withdrawn for new starts aged 22 and over from January 2026. For those learners, neither levy funds nor government co-investment can pay: the employer must fully self-fund. Funding is retained for starters aged 16 to 21, and up to 25 for care leavers and those with an Education, Health and Care Plan. For most existing staff, Levels 3 to 5 are where the funding actually is.

Who you can put through it

Existing employees qualify. Funded programmes are not just for new recruits: you can enrol people already on your payroll, and for most data and AI upskilling that's exactly who should go first, because the people who already know your business get to value fastest.

Two mechanisms govern who gets in. Recognition of Prior Learning (RPL) means funding won't pay to re-teach skills a learner already holds, so content, duration and price are reduced to reflect genuine prior experience; RPL only ever reduces, and it can't take a programme below the minimum floor. Eligibility itself turns on a simple question: after an initial assessment, does this person still have a significant amount to learn, new knowledge, new skills, new ways of working, to do the role fully well? A prior qualification, even at the same level, doesn't exclude someone who genuinely needs new skills.

The role matters more than the job title. Nobody needs a data title or a data team to qualify. People in finance, marketing, operations and line management roles do these programmes precisely because their work is full of data nobody has taught them to use. The real test is the role's day-to-day content: enough genuine data work to practise and evidence the standard's duties. What doesn't work is a role with no data in it at all; an apprenticeship can't be laid over work that never exercises the skills.

English and maths won't block an adult learner. Since 11 February 2025, learners aged 19 or over at the start don't have to pass Level 2 English and maths to complete their apprenticeship, and funded support is still there for anyone who wants to close a gap.

The under-25 economics deserve their own line, because they change what the funding can build. Take on an apprentice aged 16 to 24 and the government funds 100% of the training, whatever your levy status. Employers also pay 0% National Insurance for apprentices under 25 earning up to £50,270, and from 1 October 2026 a non-levy employer taking on a new apprentice aged 16 to 24 attracts a £2,000 hiring payment (an incentive the rules allow to be withdrawn in-year, so confirm it when you plan).

Put those together and a structured, funded graduate scheme becomes something most employers could never afford to build alone, and it widens who you can hire: degree and non-degree, younger and older, a far broader range of backgrounds. That breadth is not hypothetical for us: among learners currently on our programmes, 45% are women and 64% are from Black, Asian, mixed or other ethnic minority backgrounds (73% on our Skills Bootcamps).

And you don't need a big cohort. An organisation with one or two people joins an open cohort, training alongside peers doing the same job in other companies, which for a lone analyst is precisely the peer network their own workplace can't supply. For a closed cohort built solely around your organisation, we look for 10 to 12 people. They're different products: the open cohort buys cross-company breadth, the closed one your organisation's own context.

What funded provision asks of you, and what it gives you back

Public funding is not free money with a form attached: it's a regulated bargain, and both sides of it are worth stating plainly before you sign anything.

What you get, beyond the money, is oversight. Funded provision operates inside a regulated system: the standard publishes what must be taught and evidenced, assessment is carried out by an independent end-point assessment organisation rather than the provider marking its own work, off-the-job training must happen inside the apprentice's normal paid working hours, and providers answer to the funding body's rules and to Ofsted's inspection regime.

The bar on a funded provider is higher than on a private short course, and that's a feature: it's your assurance, and the funder's. For our part, iO-Sphere is on APAR, the register a provider must be on to deliver funded apprenticeship training, and is Ofsted-inspection-eligible.

What it asks in return is real commitment, from both employer and learner. The employer's side is protecting the off-the-job time: the published minimums on our standards run from 348 to 418 hours, which, smoothed across our 15 months of training, works out at roughly five and a half to six and a half hours a week, inside paid working hours, not evenings. The learner's side is showing up for it: attending the training, building the evidence portfolio the assessment rests on, and staying committed through the programme's milestones. There is paperwork, and there are progress reviews, because public money is being accounted for.

Say all of this to line managers and learners early. The small print of eligibility, employer backing and the working-time commitment is where funded plans actually die, and a reader who discovers it late walks away. A team that goes in with eyes open finishes.

Judging a provider before you commit the funding

Judge a funded programme by exactly the criteria you'd use if the money were leaving your own budget, because the scarcer currency, your team's working time, is leaving it either way. Four questions do most of the work:

• What will learners actually be doing, week to week? If the answer is mostly lectures and theory, that's the academic model, and we hold flatly that people get good at data and AI by doing the work, not studying it. On the programmes that use it, our learners work in Prism, a simulated e-commerce company built on 500M+ rows of real data: real data, simulated company, and the sandbox is the safety.
• Who coaches them? A coach who's done the job reads a business problem faster and teaches differently from a tutor who has taught it. Ask what the coaches have actually built.
• How will you see the return? Completion rates measure attendance, not capability. The only return that counts is changed work, and the small wins compound: automations and analyses that look trivial one at a time add up to serious hours across a year.
• How is the administration handled? RPL, evidence, reviews, pastoral care: a good provider absorbs the administration that grad schemes drown managers in.

Here is our own record, plainly. iO-Sphere has trained 900+ learners in data and AI since 2022, across bootcamps and funded qualifications, and is rated 4.8 out of 5 on Google from 78 reviews.

Where our funded delivery has had time to be measured, we publish the measurement with its basis. Our GLA-funded London bootcamp cohort ran in early 2026, and measured about four months out, 67% of its 45 starts were in employment, with 21 of those starts reaching the funder's job-outcome milestone: a rate of 47%, reported in our R12 ILR return to the GLA. We'll refresh the figures as outcomes mature.

For context on the same denominator, the DfE's positive-outcome rate for digital Skills Bootcamps nationally is 31% of starts (2023-24), measured over a longer window than our four months. London's milestone also requires the job to pay the London Living Wage, which most funding areas' definitions do not. The full side-by-side reading, every figure with its denominator, definition, checker and date, is in our bootcamp results guide.

Our apprenticeship line is newer: cohorts are still in delivery, so no completion or achievement rates exist yet to publish. The same newness is a design fact, not only a gap: the line was built after the AI shift, for the AI era, with no pre-AI curriculum retrofitted underneath it.

We're also not the right provider for everyone. If you need Level 6 or 7 standards, data science degrees, sector-regulated programmes such as FCA-aligned finance or NHS clinical data routes, or a single provider running very large simultaneous cohorts nationwide, a larger multi-sector provider will serve you better, and we say so.

Frequently asked questions

Is government-funded data and AI training actually free?

The training and assessment are covered, so it's free in cash terms for many employers, but never free in time. Levy funds and co-investment pay for approved training and the end-point assessment only; they can't pay wages, travel, kit, or the working hours learners spend on off-the-job training. Under the 2026-27 rules, apprentices aged 16 to 24 at non-levy employers are 100% government funded and you contribute 5% for those aged 25 and over. Treat the working time as the real cost, because it's where the capability actually gets built.

We don't pay the apprenticeship levy. Can we still get funded training?

Yes. If your annual pay bill is under £3m you don't pay the levy, and under the 2026-27 funding rules (in force since 1 August 2026) apprentices aged 16 to 24 are 100% government funded, with a 5% contribution for those aged 25 and over. Alternatively, a levy-paying business can transfer up to 50% of its unused funds to you, a limit raised from 25% on 22 April 2024, which can cover the full training cost.

Can we fund existing employees, or only new hires?

Existing employees qualify, and for upskilling they're usually the right people to start with. Recognition of Prior Learning reduces a programme's content, duration and price to reflect skills a learner already holds, but it doesn't remove eligibility: the test is whether the person still has a significant amount to learn to do the role fully well, not whether they already hold a qualification at some level.

Can the levy pay for short courses or one-off workshops?

Not for commercial short courses: those are paid training and should be bought as such. What has changed is that from April 2026 the Growth & Skills Levy begins funding shorter, more flexible training alongside full apprenticeships, starting with apprenticeship units from 28 April 2026: standalone units of 30 to 140 delivery hours for employed learners aged 19 and over, with no regulated qualification attached. We don't deliver units, and our view is that the small per-unit funding constrains depth; for a fast, focused need, a paid short course is often the better product.

Which data and AI standards can we fund, and at what levels?

At Levels 3 to 5, the standards we deliver are Data Technician (ST0795, Level 3), Data Analyst (ST0118, Level 4), IS Business Analyst (ST0117, Level 4, which our AI Transformation programme runs on), Data Protection & Information Governance Practitioner (ST0967, Level 4, shared by our Strategy and Governance programmes), and Data Engineer (ST1386, Level 5). Level 6 standards exist (Machine Learning Engineer, the Data Scientist integrated degree), and the AI Data Specialist sits at Level 7, where public funding was withdrawn for new starts aged 22 and over from January 2026.

How long does a funded programme take?

Months, not weeks, by design. The legal minimum for any apprenticeship is 8 months (reduced from 12 for starts from 1 August 2025), and each standard publishes a minimum number of off-the-job training hours. Our programmes run 15 months of training plus a 3-month end-point assessment, with the standard's published hours smoothed across the programme inside paid working time. Build that commitment into your planning from the start, because it's the part that decides whether the programme survives contact with the day job.

I want funded training but my employer isn't involved yet. What can I do?

You have two live routes. If you're in a job whose real work involves data, the most common path is laying an apprenticeship over the role you already hold: apply to us first, and our admissions team will help you build the business case and prepare the conversation with your employer, which lands far better than a cold, half-formed ask. If you're not employed, or your employer can't back you, Skills Bootcamps are free to unemployed and self-referred learners; ours run in funded cohort windows, so check current availability.

Ready to work out which funded route fits? We'll scope it outcomes-first with you: the capability you're buying, who it's for, and how the funding wires up. Explore our apprenticeships for employers, or if you're an employee who wants your employer to fund your training, start here and let our admissions team help you build the case.