AI for SMEs: Opportunity or Risk?

Last Wednesday I attended Annacon, a leading security conference. And as almost everywhere these days, the focus was on one subject: Artificial Intelligence. Every discussion, every deep dive revolved around AI: the promise, the risks and the impact on security.
After a day of inspiring sessions I came to one clear conclusion: AI is here, and it is here to stay. From writing tools to voice generators, from SaaS platforms to enterprise copilots, Generative AI (GenAI) is changing how companies work.
For small and medium-sized enterprises (SMEs) this shift can be a game changer or a ticking time bomb.
At Dunetrails we believe AI can help SMEs grow. But only if it is used safely and responsibly. Here are my insights after Annacon, and why this matters for every SME leader today.
The hidden AI adoption problem
Recent figures show:
- 75% of employees use AI at work.
- 78% use their own AI accounts.
- 52% do not admit that they use AI.
That means company data can already end up in uncontrolled environments, without you knowing. And it is not just personal data, but sensitive business information too.
I have seen employees e-mail confidential files to their private account to run them through ChatGPT because the company firewall blocked access. I have seen people secretly record meetings on their phone to have them processed by dubious transcription bots. People are resourceful, and if you do not give them safe tools, they will find unsafe ones.
What does safe AI adoption look like?
At Dunetrails we define safe AI adoption in three steps:
Awareness Employees need to understand the risks. Public AI chatbots do not just answer questions, they collect data. Some are outright malicious, designed to harvest information.
Governance Apply the principle of least privilege. Just as with IT security, employees should only have access to the data they need. Tools such as Microsoft Purview help enforce this.
Offer safe alternatives If you do not provide a company-controlled AI solution, employees will look elsewhere. Microsoft Copilot is a good example of a safe, integrated option that keeps data within your trusted environment.
Innovation vs. compliance: finding the balance
We are in favour of innovation, but not at the expense of ethics or security. Regulation such as NIS2 has brought awareness and better practices, even if it required investment. Compliance is not a checkbox; it is a foundation for trust.
At the same time, innovation drives progress. SMEs have an advantage here: shorter decision cycles mean faster adoption and agility. Those who embrace AI smartly will thrive. Those who ignore it, or jump in blindly without guardrails, will fall behind or worse.
My biggest frustration
Companies that embrace AI head over heels without understanding the basics. AI is an amplifier, not a replacement. Do you have no software development experience and are you using AI to build an app? Expect gaps: security gaps, compliance problems and operational risks. But if you are an experienced developer, AI makes you faster and better because you know how to steer it.
The same goes for every domain. AI amplifies your strengths, but also your weaknesses.
The Dunetrails promise
We help SMEs turn AI into an opportunity without sacrificing security. Our approach:
- Education: make your people aware of risks and best practices.
- Governance: implement policies and tools such as Purview.
- Safe AI: use trusted solutions such as Microsoft Copilot to keep data safe.
One piece of advice for 2025
Inform yourself and your people, and give them access to AI they can use safely under company control. Because if you do not manage AI in your company, someone else will.
Closing thought
AI will reshape every industry. For SMEs the question is not if but how. Safe adoption is not an option; it is the only way forward.



