Insight from 100+ Scottish Organisations About Adopting AI Securely
Our recent webinar with Microsoft specialists Ben Voce and David Fortnam drew in more than 100 people from Scottish accountancy firms, housing associations, charities, professional services businesses, and public sector bodies. Everyone came with the same underlying question: how do we adopt AI in a way that is safe, well governed, and actually delivers value?
Across six live polls and a lively hour of conversation, we heard something that we suspect a lot of leadership teams need to hear. AI is already here. The question is no longer whether to adopt it, but how to do it well.
Watch the webinar
If you missed the session, or want to share it internally with your leadership team, IT lead, or wider colleagues, you can watch it in full below.
The AI conversation has moved on
For years, the debate in most organisations has been whether AI is something to worry about later. That debate is over. AI is already inside almost every business, whether it has been formally approved or not.
The first poll of the session confirmed it. When we asked people where they were on their AI journey, only a small minority said they had not started at all.
Six in ten organisations described themselves as “dabbling”. That is a huge amount of experimentation happening without a plan, without training, and often without the leadership team knowing quite what is going on. It is not a bad place to start, but it is a risky place to stay.
Dabbling means individuals are trying things out on their own. Rolling out means the organisation is doing it together. The difference matters because it is the second group who will pull ahead over the next 12 months, not because they have more curiosity, but because they have more structure.
Shadow AI is not a future risk – it is already happening
When we asked people which AI tools they had personally been trying at work, the answer was much broader than a single platform. Microsoft Copilot was on top by a clear margin, but ChatGPT was used almost as widely, and Claude, Gemini, and other tools were all quietly making appearances.
The important point is not which tool “wins”. It is that most people are using more than one, sometimes on personal accounts, sometimes with real work data, and often without their organisation knowing.
That is exactly what the industry has started calling shadow AI, and the concerns raised in the session were just as consistent as the usage patterns.
When you put the two polls side by side, the picture is clear. People are using tools, worried about the risks of using tools, and often not sure who else is using what. That is not a healthy place for any organisation to sit for long, particularly one handling client, financial, or personal data.
The most encouraging thing about the concerns raised is that all of them are solvable. Not with a single silver bullet, but with sensible policies, a clear position on which tools are approved, better SharePoint hygiene, sensitivity labels, and grown-up conversations about how AI is used across the business.
Claude or Copilot? That may no longer be the right question
One of the most talked-about themes during the session was the growing interest in Claude, particularly across accountancy and legal firms. A fair chunk of the audience are already trying it, and some are comparing it to Copilot as though they are alternatives.
Microsoft has now brought Anthropic’s Claude models inside Microsoft 365 Copilot for eligible organisations, subject to admin settings and region. For most organisations, this changes the shape of the conversation entirely.
As Ben Voce put it during the live Q&A:
Claude is now part of Copilot, but the main reason to use Copilot is that it sits within the Microsoft applications, so you do not need to jump between apps.
In other words, the choice is not simply about which model is smartest today. It is about where AI is being used and how it is controlled. When AI runs inside the tools your team already work in – Outlook, Teams, Word, Excel, PowerPoint, and SharePoint – access, permissions, security, and reporting can all be managed through your Microsoft tenant. That is a very different conversation to standalone AI tools being used through personal browser tabs.
Buying the licence is not the same as getting value
One of the most striking results of the entire session came when we asked where organisations were with Microsoft 365 Copilot specifically.
Only a small minority of organisations with Copilot licences feel they are getting good value from them. The rest are paying for a powerful tool that is quietly under-used.
This is not a Copilot problem, and it is not a failure of the technology. It is an adoption problem. Buying the licence is the easy bit. The value comes from making Copilot part of how people actually work – from meeting recaps and inbox triage, to first drafts, summaries, analysis, presentations, and repetitive tasks that no one enjoys doing.
Organisations that see the biggest return from Copilot tend to have three things in common. They give people permission to experiment safely. They build small, practical training and habits into everyday work. And they treat AI adoption as a leadership responsibility, not an IT project.
Governance is not a blocker – it is the runway
One of the quieter moments of the session came when we asked how confident people felt about their data governance being ready for AI. The results speak for themselves.
More than nine in ten respondents either know they have work to do, are unsure, or are outright worried about how ready their data foundations are. Only five percent felt fully confident that their policies, labels, and controls are in place.
This matters because of how Copilot works. It respects the permissions your organisation already has. It surfaces information a user is allowed to see. If access rights, sensitivity labels, and SharePoint structure are not in good shape today, Copilot will not fix that. In fact, it will make it easier for people to stumble across information they were never meant to see – faster.
Good governance is not a blocker to AI. It is the runway that lets AI take off safely.
What organisations really want next
The final poll of the session asked what would help most in moving forward with Copilot. The answers were refreshingly practical.
Notice what is not on this list. No one asked for another AI tool. No one asked for more slide decks about the future. What organisations are asking for is time, training, structure, and confidence.
These are the essentials of any successful technology rollout, and they map neatly onto the three pillars Microsoft themselves talk about for Copilot success – leadership commitment, human change, and technical readiness. Leadership sets the direction. People need help to change how they work. And the technical foundations must be in place to support both.
The Lugo view
What we heard on the day, and what the poll data quietly confirms, is that Scottish organisations are ready to move forward with AI. What they want is a sensible path, not a sales pitch.
Adopting AI securely and confidently is not a one-off event. It is an ongoing programme that touches licensing, security, governance, training, culture, and communication. For regulated Scottish organisations – accountancy firms, housing associations, wealth management businesses, and charities – there is no shortcut around getting the foundations right. But nor is there any excuse for delay. The people in your team are already using AI. The question is whether they are doing it inside a safe, well-governed environment or outside it.
The organisations that will get the most from AI over the next 12 months are the ones that combine four things:
- Leading – clear leadership sponsorship and a real AI strategy
- Uniting – bringing people, tools, and processes together securely
- Growing – enabling teams to work in new ways with confidence
- Outstanding – doing it all with integrity, transparency, and control
Those are the values Lugo was built on, and they are exactly the values that guide how we help our clients adopt AI well.
How Lugo can help you move forward
Based on what organisations told us they need, Lugo can support your firm with:
- A Copilot readiness assessment as part of our Technology Alignment process, reviewing identity, access, permissions, SharePoint hygiene, and sensitivity labelling
- Adoption and training support for your team, including prompt-a-thons and role-based sessions
- Help drafting or refining your internal AI policy in line with UK GDPR and sector expectations
- Licensing, security, and ongoing management of Microsoft 365 Copilot through a trusted partner
- Access to the Lugo Copilot User Group, a peer community for Lugo clients with Copilot licences to share what works, what does not, and where AI is heading next
Download the slides
You can download the full Microsoft 365 Copilot Inspire slide deck used in the session below. It is a useful resource to share with your leadership team, your IT lead, and anyone across the business who is thinking about how to bring Copilot into everyday work.
Ready to have a proper conversation about AI?
If you would like to talk about where your organisation is on its AI journey and where Copilot could fit in, we would love to hear from you. Whether you are dabbling, rolling out, or somewhere in between, Lugo can help you take the next step securely and confidently.
