A thought leadership article for SMB and mid-market organisations preparing for practical, secure AI adoption.
Figure 1: The pendulum of computing
During my time at Dell/EMC as Head of Programme Office in EMEA for the Cloud Service Provider Program, I worked with partners and organisations navigating the early evolution of hyperscale cloud. From around 2010 to 2016, much of the market narrative suggested that large-scale public cloud would become the natural destination for almost every workload.
The reality was more balanced. Many organisations wanted the scale and flexibility of cloud, but they also needed control. Their core company data, legacy applications and business-critical processes were not always easy to move. Some workloads were too sensitive. Some were too integrated. Some were too important to risk disrupting. That is why private cloud, hosted private cloud and hybrid cloud became such important models.
Now AI is creating a similar moment. Cloud AI will be an important starting point for many SMBs and mid-market organisations. But as AI moves from experimentation into everyday operations, businesses will need to think carefully about where their intelligence runs, where their data lives, and how they keep control of cost, security and governance.
|
AI readiness starts with data readiness, and data readiness starts with infrastructure. |
Computing has always moved between centralisation and control. In the 1970s and 1980s, mainframes concentrated processing power in the data centre. In the 1990s, file servers and local networks brought business data closer to departments and users. Through the 2000s, client-server computing and virtualisation created more distributed, flexible architectures.
The cloud era accelerated that shift again. Public cloud services began maturing after the launch of services such as Amazon S3 and EC2 in 2006, while the early 2010s saw private cloud and virtualisation-led infrastructure become a serious option for organisations that needed efficiency and control. By the mid-to-late 2010s, hybrid cloud became the practical compromise: some workloads in public cloud, some in private environments, and many businesses learning that one model did not fit every use case.
The 2020s then pushed many organisations into a more cloud-first mindset, helped by remote work, SaaS adoption and the wider digital transformation agenda. The UK Government introduced its Cloud First policy in 2013, and the cloud-first idea became a mainstream default across many sectors in the years that followed.
AI is now causing another rebalancing. The future is unlikely to be all-cloud or all-on-premises. It is more likely to be hybrid: public cloud for reach and specialist services, private cloud for sensitive or repeatable workloads, and AI-enabled endpoints for local productivity.
AI is not just another application. It is a new compute and data layer. It can search, summarise, generate, automate, classify and recommend. It can work across emails, documents, files, customer records, finance systems, service histories, policies and operational knowledge. That makes it powerful, but it also makes the infrastructure question more important.
For SMBs, the challenge is not simply whether they can access AI tools. It is whether they can use AI safely, affordably and in a way that supports the business rather than creating new risk. As adoption grows, cloud AI costs can become harder to predict. Usage-based services, token consumption, AI agents, automation, data movement and repeated inference can all add up.
That does not mean public cloud AI is the wrong answer. It means SMBs need a workload strategy. Some use cases will belong in the cloud. Some will run locally on AI-enabled PCs. Some business-critical or data-sensitive workloads may be better suited to private or hybrid infrastructure.
Small and mid-sized businesses are the heartbeat of the UK economy. The Federation of Small Businesses reports that SMEs account for 99.9% of the UK business population, around three-fifths of private sector employment and roughly half of private sector turnover. If AI becomes a productivity advantage, SMBs cannot afford to stand still.
The risk is that AI adoption splits the market. Larger organisations have the budget, skills and infrastructure to experiment widely. SMBs may start with public AI tools and productivity assistants, but could then face three difficult outcomes: they get left behind because AI feels too complex, they face unpredictable cloud processing costs as usage scales, or they use AI in ways that expose sensitive data because the right controls are not in place.
The better answer is not to avoid AI. It is to prepare for AI properly. That preparation starts with modern data infrastructure: secure, resilient, well-governed, right-sized and able to support cloud, private cloud, edge and endpoint workloads.
Figure 2: SMBs are central to UK growth, which makes practical and affordable AI adoption a business-critical issue.
Before AI is scaled across a business, the security foundations need to be right. Cyber Essentials gives SMBs a practical baseline. The National Cyber Security Centre describes Cyber Essentials as the minimum standard of cyber security recommended by the UK Government for organisations of all sizes, aligned to five technical controls designed to prevent the most common internet-based cyber security threats.
Those five controls matter even more in the AI era: firewalls, secure configuration, user access control, malware protection and security update management. Put simply, businesses need to make sure the doors are locked before AI is given access to more systems, more users and more data.
But AI readiness must go beyond certification. It should also include endpoint protection, identity and access management, backup and cyber recovery, user awareness, approved AI tooling, and clear policies on what data can and cannot be used in AI platforms. This is especially important as AI becomes embedded into laptops, desktops, collaboration tools, private cloud platforms and public cloud services.
Figure 3: Cyber Essentials is a strong starting point, but AI readiness also requires governance across devices, cloud, data and people.
AI makes data governance impossible to ignore. In the cloud-first era, many SMBs focused on moving applications and data to platforms that were easy to consume. In the AI era, the question changes. Businesses must now decide which data AI can access, where that data should live, who is allowed to use it and how sensitive information is protected.
A practical AI policy should define approved tools, rules for sensitive data, access controls, human review requirements, where different AI workloads should run, and what happens if data is accidentally exposed. This is not bureaucracy for its own sake. It is how an organisation turns data into trusted intelligence rather than uncontrolled risk.
Dell Technologies research underlines the challenge. In the Innovation Accelerated eBook, 69% of organisations said they struggle to get value from data, while only 26% anchor all innovation efforts in data insights. The same paper highlights the need to seek smart, be flexible and build resilience: a useful framework for SMBs planning their AI foundations.
An AI-ready SMB foundation is not one product. It is an architecture and a set of controls that work together. It starts with modern endpoints, including AI-ready PCs that can support local productivity features and allow some AI inference and productivity functions to run locally rather than in the cloud. AI-capable PCs are rapidly becoming mainstream. Gartner forecast in 2025 that AI PCs would represent over 50% of the global PC market in 2026 although industry memory shortages could cause the adoption rate to slow a little.
It then extends into identity, access control, data protection, backup, recovery, storage, cloud connectivity and private or hybrid cloud options. For some workloads, public cloud will remain the right choice. For others, particularly where data sovereignty, predictable workload economics, latency, regulatory requirements or greater infrastructure control are important; private or hybrid infrastructure may be more appropriate.
This is where the modernising data infrastructure conversation becomes relevant to SMBs. It is not an abstract enterprise concept. It is about making sure the business can access the right data, protect it properly, recover it quickly and use it to support analytics, automation and AI.
The most important starting point is not to ask, “Which AI tool should we buy?” It is to ask, “Are our data, devices, security controls and infrastructure ready for AI?”
Deane Computer Solutions, working with Dell Technologies, can support SMB and mid-market organisations consultatively through that journey. That may include Cyber Essentials support, reviewing data governance and access controls, modernising endpoint devices, strengthening backup and cyber resilience, assessing cloud and on-premises workloads, and helping customers plan practical private or hybrid AI foundations.
For many organisations, the right first step will be a readiness conversation: what data matters most, where it lives today, what needs protecting, where AI could add value, and which workloads should remain in the cloud, move closer to the business or run locally on AI-enabled devices.
|
AI may be the future — and it increasingly looks like it will be — but infrastructure will decide how safely, affordably and successfully mid-market and SMB organisations can use it. |
Reach out to Deane Computer Solutions to explore how we can help you make the transition in IT.
If you are planning to adopt Microsoft Copilot, refresh workplace devices, strengthen Cyber Essentials, modernise data infrastructure or explore private and hybrid AI options, Deane Computer Solutions can help you assess where you are today and what needs to happen next.
Ask about our Secure Copilot 30 Day Sprint program.
Call: +44 1305 250505 | Email: sales@deane247.com
• Dell Technologies, Innovation Accelerated: Power innovation with modern data infrastructure, pages 5, 8 and 12.
• Federation of Small Businesses, UK Small Business Statistics, 2025.
• National Cyber Security Centre, Cyber Essentials overview and Cyber Essentials requirements for IT infrastructure.
• Amazon Web Services, Our Origins, citing the 2006 launch of S3 and EC2.
• GOV.UK, Government Cloud First policy, introduced in 2013.
• Gartner, Gartner Forecasts Worldwide Shipments of AI PCs to Account for 43% of All PCs in 2025.