°Ç¸ç³Ô¹Ï

About the AI Skills BoostÌýprogramme

AI Skills Boost is a government-industryÌýinitiativeÌýto improve UK workforce readiness by upskilling 10 million UK workers in AI skills by 2030, which was announced by the Prime Minister in June 2025 at London Tech Week. The programme focuses onÌýleveraging theÌýreach andÌýhigh quality training provisionÌýofÌýgovernment andÌýindustry partners toÌýequip the UK workforce withÌýAI skillsÌýto support a range of outcomes, including:

• supporting workers to adaptÌýand thrive in a workplace where AI tools are increasinglyÌýwidespread
• targeting support towards worker groups and cohortsÌýthatÌýmost requireÌýit
• facilitatingÌýwidespread uptake of AI across the economy to support productivity and growth

What has the programme achieved so far?

1,001,147ÌýAIÌýtraining courses have been completedÌýaccording to courseÌýcompletion data shared with DSIT by industry partnersÌýinÌýJanuaryÌý2026.ÌýThis figure covers all AI skills courses delivered by partners since June 2025, from introductory AI literacy to advanced training in areas like data science and machine learning engineering, for both external learners (customers, clients, platform users) and internal learners (partner employees).ÌýThe 11 industry partners involved inÌýdelivering these coursesÌýconsist ofÌýAccenture,ÌýAmazon, BT, Barclays, IBM, Google, Intuit, Microsoft, SAS, Sage,ÌýandÌýSalesforce, and theÌýoverallÌýcourse completion figure includes courses delivered through the government’s One Big Thing initiativeÌýin 2025.ÌýSpecific partner or course-level breakdowns of course completion are not shareable due to commercial sensitivity.Ìý

What is the potential economic impact of the programme?Ìý

TheÌýprogramme has significant potential to make upskilled workers more productive and increase earnings. For example,Ìýresearch fromÌýthe IMF indicates that AI skills in the UK are associated with wage premiums of about 7.5–8 percent within occupations,^{[footnote 1]} and more broadly, AI use is associated with substantial productivity gains.^{[footnote 2]}ÌýIf we cautiously assume that the training can unlock an additional 1% in output for the upskilled workers,ÌýandÌýthe ambition to trainÌý10 million workers is met,Ìýthis would be equivalent to over £7.5 billionÌýhigherÌýGVA per annum.ÌýThis is calculated by multiplying average worker GVA (~£77,000 inÌý2024)^{[footnote 3]}Ìýby theÌýnumber of workers in scopeÌýof the programme by 2030Ìý(10 million), andÌýmodelling aÌý1%Ìýincrease inÌývalue added.ÌýHowever, this figure is highly uncertain, and we willÌýmonitorÌýthe programme toÌýmore accurately measure economic returns.Ìý

What is the potential impact of AI on UKÌýproductivity?Ìý

The OECD estimates that AI adoption could add 0.4 to 1.3 percentage points to the UK’sÌýannual labourÌýproductivity growthÌýover the next decade.^{[footnote 4]} The range in productivity uplift is informed by the adoption rate scenario (based on historic technology adoption rates) and whether complimentary software is developed (no significant uplift in AIÌýcapabilitiesÌýis assumed).ÌýÌý

DSIT hasÌýundertakenÌýadditionalÌýanalysis to model this impact in GVA terms

1. Firstly, we model an adoption curveÌýfor each adoption scenario (slow, medium and rapid), using a constantÌýmultiplyingÌýfactor that provides a compoundingÌýadoptionÌýcurveÌýfrom initial 2024 adoption in the paper (5%) to adoption in 2034 (30%, 47%,Ìý60%),^{[footnote 5]}ÌýseeÌýworked example below.
   

2. Next, we use the adoption rate calculated above as an adoption adjustment. WeÌýfind thisÌýbyÌýdividingÌýthe adoptionÌýrateÌýinÌýa givenÌýyearÌýby the adoption in the final year (Equation 2).
   

3. Thirdly, using the OECD’s estimated improved annual average productivityÌýgrowthÌýratesÌý(0.39, 0.97,Ìý1.27), we calculate the total productivity uplift in 10 years, reachingÌý4.0pp,Ìý10.1pp and 13.5pp in 2034 (see Equation 3).
   

4. We then apply the adoption adjustment toÌýthe totalÌýproductivity uplift to obtain an adoption-adjusted productivity uplift inÌýeachÌýyearÌý(see Equation 4).
   ÌýÌý

5. AsÌýa final step, the percentage point uplift calculated above is applied to forecasted GVA (developed usingÌýMarch 2025ÌýOBR GDP growth forecasts^{[footnote 6]}Ìý²¹²Ô»åÌý2024ÌýGVA^{[footnote 7]}). The calculated difference in GVA constitutes theÌýproductivityÌýbenefits of AI adoption.
   ÌýÌý

We find theÌýestimatedÌýproductivityÌýgrowth is equivalentÌýtoÌýaddingÌýbetweenÌý£55-140 billionÌýto UK GVA in 2030.ÌýA worked example for the medium scenario is show below.ÌýThisÌýestimate is highly uncertain, and DSIT will updateÌýthis analysis as new evidenceÌýemerges.Ìý