A new report effort in Wales is putting a hard question at the center of the country’s technology debate. Can Wales turn AI from something its firms buy into something its economy builds? Business News Wales reported that its Artificial Intelligence, Technology and Digital Economy Editorial Board, created as part of the INFORM Wales initiative, has examined how Wales can build a stronger digital economy. The board’s findings are meant to inform an INFORM evidence report on AI, technology and the digital economy in Wales.
One caveat belongs up front. The exact headline “New AI Report to Challenge How Wales Can Turn Technology Into Economic Opportunity” does not map cleanly to one document in the sources we reviewed. Three separate pieces of work sit around the story: the INFORM evidence report still to come, NatWest Group’s AI Adoption Report, and an earlier AI in Wales report tied to Tom Giffard. This analysis treats them together and labels which fact comes from which. For the wider debate about whether AI’s economic prize survives rising costs, see our look at McKinsey’s AI value curve.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What Business News Wales Reported
The core facts are short and specific. Business News Wales said the editorial board belongs to INFORM Wales, and that its discussion will feed an evidence report. That report will list issues raised by participants, flag areas that need further investigation, and pose practical questions for industry and government. In other words, the document is built to provoke the next round of work rather than close the debate.
The framing matters. A report that hands over a list of questions is a different instrument from one that hands over a policy blueprint. It asks business leaders and ministers to answer in public, and it gives journalists and researchers a checklist to hold them to. For a small economy with limited room to fund every ambition, that approach can sharpen priorities. It can also stall if nobody owns the answers.
The board’s central prompt looked ahead. According to the report, participants asked what Wales must do over the next five years to speed up AI and digital adoption and become a more competitive digital economy. Five years is long enough to cover a full cycle of skills training and infrastructure build-out. It is also short compared with how quickly the AI product market now changes.
INFORM Wales and the Editorial Board Explained
INFORM Wales is the umbrella initiative. The Artificial Intelligence, Technology and Digital Economy Editorial Board is one of its working groups, and it exists to gather views from industry and government in one room. Business News Wales presented the board as a listening exercise with a defined output, the INFORM evidence report.
That structure gives the exercise a few useful traits. It brings named sectors into one conversation, which tends to surface conflicts early. A telecoms provider worries about physical infrastructure, a bank worries about trust, and a university worries about graduate pipelines. When those worries land in the same document, readers can see which constraints repeat. Repetition is the best signal a policy team has for where to act first.
The structure also carries a risk. Editorial boards produce summaries of opinion, and opinion is not measurement. The planned report promises to identify areas that need further investigation, which is an honest admission of that limit. Readers should treat the first version as a map of questions, not a scoreboard of results.
The Six Areas the Board Examined
Business News Wales listed the territory the board covered. It spans digital and physical infrastructure, technology adoption by Welsh businesses, cyber resilience, leadership, workforce skills, and the conditions that let technology companies start, scale and invest in Wales. Together these form a full chain from cables and data centers to boardroom decisions to hiring.
Infrastructure and cyber resilience
Infrastructure comes first because every AI tool depends on connectivity and compute somewhere. Cyber resilience travels with it. A firm that moves customer data into AI workflows widens its attack surface, so security cannot trail adoption by a year. Boards that treat the two as separate budgets usually pay for the gap later.
Leadership, skills and company formation
Leadership and skills are the human half of the list. Executives decide whether AI stays a pilot or reaches payroll, sales and operations. Skills decide whether staff can use it well. The last item, conditions for firms to start, scale and invest, asks the sharpest question of all: why would a technology company choose Wales over somewhere else?
| Board topic | Question it raises | Evidence worth watching |
|---|---|---|
| Digital and physical infrastructure | Can firms reach the connectivity and compute that AI tools need? | Named gaps and who funds them |
| Business technology adoption | Are firms moving from trials to daily use? | Survey base sizes and sector splits |
| Cyber resilience | Does security keep pace with adoption? | Incident data and training offers |
| Leadership | Do executives own AI decisions? | Board-level plans, not pilots |
| Workforce skills | Who trains staff to use and supervise AI? | Course capacity and employer uptake |
| Start, scale and invest | Why would a tech company choose Wales? | Funding routes and talent supply |
The third column in that table is our editorial reading, not content from the board. It lists what a careful reader should look for when the INFORM report arrives.
NatWest’s AI Adoption Report: 80%, 73%, 59%
The hardest numbers in the Wales conversation come from banking, not government. NatWest Group’s AI Adoption Report states that, among Welsh businesses already using AI, 80% reported improved efficiency, 73% reported increased innovation and 59% reported revenue growth. The press release headline adds a second message: trust can take those gains further.
Those three figures describe three different kinds of payoff. Efficiency is the easiest to claim because it lives inside the firm and shows up in hours saved. Innovation is harder, since it means new products, services or ways of working. Revenue growth is the hardest, because customers have to pay for the change. Seeing 59% report revenue growth is the most commercially interesting number in the set.
| Outcome reported by Welsh AI-using firms | Share reporting it | What kind of gain it signals |
|---|---|---|
| Improved efficiency | 80% | Cost and time savings inside the firm |
| Increased innovation | 73% | New products, services or processes |
| Revenue growth | 59% | Customer-facing commercial return |
Source for all three percentages: NatWest Group’s AI Adoption Report. The base is Welsh businesses that already use AI, not all Welsh businesses.
Reading the NatWest Numbers Carefully
Good headlines hide a denominator. The 80%, 73% and 59% figures apply to firms that already adopted AI. That group likely skews toward businesses with the skills, money and appetite to try it. A firm that struggled with AI and quietly dropped it may not be in the sample at all. Readers should avoid stretching these percentages to the whole Welsh economy.
We could not confirm sample size, survey dates or sector splits from the material available to us, so we do not state them. Those details decide how far the findings travel. A survey of a few hundred firms in one quarter tells a different story from a multi-year panel. The INFORM report is the natural place to ask for that detail.
Self-reported gains also differ from audited ones. Managers who adopted AI have reasons to say it worked. That does not make the results false, but it does argue for pairing survey data with hard measures such as productivity statistics. The trust theme in NatWest’s headline connects here. Firms that cannot explain what an AI tool did, or who is accountable when it errs, will hesitate. Our coverage of shadow AI use among workers shows the flip side, where staff adopt tools faster than their employers govern them.
Tom Giffard’s AI in Wales Report: 50 Companies, 40 Investors, 5 Hubs
The third document predates the other two. Source Security reported in May 2024 that Tom Giffard, a member of the Welsh Parliament, worked with Deep Knowledge Group and the Global AI Industry Ecosystem Association on an AI in Wales platform and report. It profiles 50 companies, 40 investors and 5 hubs.
That is a mapping exercise. It answers who is here, who funds them and where they cluster. Mapping reports are valuable at the start of an ecosystem’s life because investors and founders cannot connect with players they cannot see. They are less useful for judging outcomes, because a directory lists activity without measuring results.
Giffard framed the report in positive but measured terms. He said the document outlines the potential advantages of AI in the Welsh economy and the opportunities a strong AI ecosystem could offer, with encouraging analysis of the current landscape while carefully considering future challenges. The tone is optimistic with a hedge, which fits a directory-style report.
Three Documents, One Conversation
Putting the three side by side shows how they complement one another. None of them answers the whole question alone. Each covers a different layer of the problem, and the gaps between them are where the next round of research should aim.
| Document | Type | Key figure or scope | Main limit |
|---|---|---|---|
| INFORM evidence report (forthcoming) | Editorial board findings | Issues, research gaps and practical questions on infrastructure, adoption, cyber, leadership, skills and company growth | Opinion-led, questions not answers |
| NatWest Group AI Adoption Report | Business survey | 80% efficiency, 73% innovation, 59% revenue growth among AI users | Covers firms already using AI |
| AI in Wales (Giffard, Deep Knowledge Group) | Ecosystem mapping | 50 companies, 40 investors, 5 hubs | Directory, not an outcomes measure |
Read together, they suggest a sequence. The 2024 mapping shows the supply side. The NatWest survey shows demand and early returns. The INFORM report tries to connect both to decisions. A reader who wants a fast read should start with the NatWest figures, then check the INFORM questions against the gaps listed above.
Consume or Build: The Capability Question
The most interesting thread in the Business News Wales account is the distinction between using technology and building it. The board discussed how Wales could develop more of the capability, expertise and economic value tied to creating and deploying technology, instead of only consuming what others make. That is a strategic claim, and it carries real tradeoffs.
Consuming is cheap and fast. A Welsh firm can subscribe to a frontier model and gain efficiency within weeks. The value, however, flows mostly to the provider. Building is slower and riskier, yet it keeps skills, intellectual property and higher-paid jobs in the country. Most regions land on a mix: consume the general-purpose layer, build on top of it in sectors where local expertise is deep.
The speed of the model market sharpens this choice. Releases arrive in clusters, with new tiers, context windows and prices appearing within weeks of each other. Our report on Gemini 4 Argon and its 1M-token paid plans is one example of how quickly a firm’s tool menu can shift. A Welsh business that hard-wires one vendor into its workflow risks repricing or retirement of that product. Flexible architecture and trained staff protect against that churn better than any single platform bet.
What Participants Said
Business News Wales quoted board participants without naming the speakers in the passages we reviewed, so we attribute the words to the board and the outlet. One participant said: “AI and digital transformation are going to affect virtually every part of the Welsh economy, but the important conversation isn’t simply about the technology itself. It’s about whether Wales is creating the conditions that allow businesses to actually use it, develop it and benefit from it.”
Another described the group’s purpose this way: “Our role is to listen, interrogate the evidence and turn those conversations into useful intelligence that can inform businesses and the people making decisions about Wales’ economic future.” Both statements come from the Business News Wales account of the board’s discussion.
The first quote reframes the debate. It treats the technology as the easy part and the conditions around it as the hard part. That echoes the board’s list of infrastructure, skills, cyber resilience and investment climate. The second quote sets expectations for the report itself: intelligence for decision-makers, built from evidence, rather than advocacy.
Tom Giffard’s earlier line, covered above, completes the set. All three voices stress opportunity paired with conditions. Nobody in the material we reviewed argued that adoption would happen on its own.
Market Impact for Welsh Businesses and Investors
For Welsh firms, the near-term impact is practical. If 59% of AI-using firms report revenue growth, competitors who have not started face a growing gap. Businesses outside that group should expect customers, suppliers and lenders to ask what their AI plans are. The questions will arrive from banks first, since NatWest has already put the topic on its agenda.
For investors, the board’s last topic is the one to watch. Conditions for technology companies to start, scale and invest decide whether capital flows to Wales or elsewhere. The AI in Wales mapping of 40 investors and 5 hubs shows a starting base. Whether the INFORM report names concrete barriers, such as follow-on funding or talent supply, will shape how investors read it.
Workers sit in the middle of this. Skills sit on the board’s list for a reason, and the global labor picture is uneven. Our piece on Goldman Sachs and AI hiring headwinds shows how fast employers can change hiring plans once AI tools prove their worth. Training capacity in Wales will determine whether local workers gain from adoption or watch the gains go to imported talent and imported software.
Historical Context: From Mapping to Evidence
The Welsh conversation has moved through recognizable stages. In 2024 the AI in Wales project catalogued who was operating in the ecosystem. In 2026 the emphasis shifted toward measured outcomes, with NatWest publishing adoption results and the INFORM board preparing an evidence report. That progression mirrors how many regions treat a new technology: first awareness, then adoption surveys, then policy questions about capability and competitiveness.
The stage matters because each one demands different proof. Awareness needs directories. Adoption needs survey data. Competitiveness needs comparisons with other places, which none of the three Welsh documents appears to supply on its own in the material we reviewed. That is the largest gap we see, and the INFORM report could fill it by naming comparable regions and the measures used.
For readers following the broader AI landscape, our AI and machine learning coverage tracks the model releases and funding rounds that set the backdrop for local debates like this one.
How Wales’ Approach Compares
Wales is not alone in asking how a mid-sized economy captures AI value. Governments are taking different routes, and two recent items on this site show the contrast. The WAICO and Pax Silica blocs show countries choosing sides in a geopolitical contest over AI. The UK’s £100M biosecurity lab with OpenAI shows a national government backing a specific, high-value use case with named industry partners.
Against those, the Welsh approach is bottom-up and consultative. It gathers business and government views, then publishes questions. That costs less and moves fewer politically sensitive levers, though it also commits no money. A region that wants to build capability, not just consume it, will eventually need a funded program tied to a specific sector. Candidates the board’s own list points toward include cyber resilience and workforce training, since both appear in its scope.
Five Predictions for Wales’ AI Agenda
These are Tech Insider’s expectations, not claims from the sources. Each rests on the facts above and could change as the INFORM report appears.
- The INFORM report will lead with skills and adoption support. Both appear in the board’s scope and both match the NatWest finding that benefits are real but uneven.
- Cyber resilience will move from side topic to gating issue. Firms adopting AI widen their exposure, and the board listed resilience as its own area.
- Welsh firms will favor vendor-flexible setups. With frontier models changing tiers and prices within weeks, locking into one provider looks riskier each month.
- Trust will become a selling point. NatWest’s own headline links trust to further gains, so expect lenders and suppliers to ask for governance evidence.
- The next survey wave will face scrutiny on method. The 59% revenue figure is the number most likely to be challenged, so sample size and sector mix will matter.
The common thread is that evidence quality will decide credibility. A Wales that publishes clear methods and honest limits will earn more investor attention than one that publishes bigger headline percentages.
What to Watch Next
Three signals will show whether this effort changes anything. First, whether the INFORM evidence report names owners for its practical questions, in industry and in government. Questions without owners tend to fade. Second, whether it adds comparison data from other regions, which would turn a local discussion into a competitiveness test. Third, whether business groups and banks publish follow-up adoption data using consistent definitions, so year-over-year change becomes visible.
Readers should also keep the headline caveat in mind. Until Business News Wales or the board confirms which document the phrase “new AI report” refers to, treat the three pieces as a connected body of work. That keeps expectations realistic and avoids crediting one report with claims that belong to another.
FAQ: AI in Wales
What is the INFORM Wales Artificial Intelligence, Technology and Digital Economy Editorial Board?
Business News Wales reported that the board was created as part of the INFORM Wales initiative. Its discussions are meant to inform an INFORM evidence report on AI, technology and the digital economy in Wales.
What will the planned evidence report cover?
It will identify issues raised by participants, areas that need further investigation, and practical questions for industry and government. The board examined infrastructure, business adoption, cyber resilience, leadership, workforce skills, and conditions for technology firms to start, scale and invest.
What did NatWest find about AI use in Wales?
NatWest Group’s AI Adoption Report says that among Welsh businesses already using AI, 80% reported improved efficiency, 73% reported increased innovation and 59% reported revenue growth.
Do the NatWest figures apply to all Welsh businesses?
No. The percentages describe firms that already use AI. They do not measure businesses that have not adopted it or that tried and stopped.
What is the AI in Wales report?
Tom Giffard worked with Deep Knowledge Group and the Global AI Industry Ecosystem Association on an AI in Wales platform and report. It profiles 50 companies, 40 investors and 5 hubs, according to Source Security.
Is the headline report the NatWest report, the INFORM report or the AI in Wales report?
The sources we reviewed do not establish that. The headline exists in the story’s framing, but the available reporting does not tie it to one specific document.
Why does the consume-versus-build question matter?
Business News Wales said the board discussed building more capability, expertise and economic value in Wales rather than only consuming technology developed elsewhere. Building keeps skills and higher-value work in the country, while consuming delivers faster but lets most value flow to foreign providers.
What timeline did the board consider?
The discussion asked what Wales needs to do over the next five years to accelerate AI and digital adoption and become a more competitive digital economy.
