Lucid Computer Solutions Services Logo
PHP Critical Website Update Guidance
22nd October 2026
SharePoint 1TB Storage Limit
6th November 2026
Show all

Small businesses should be aware of AI-based cyber threats, but there is no need to panic. AI mainly helps criminals do what they already did, only better: writing convincing phishing emails without spelling mistakes, cloning voices, and producing believable fake invoices. The attacks still depend on someone clicking a link, paying a bill, or handing over a login, so the proven basics still work. Multi-factor authentication, advanced email filtering, a firm process for checking payment requests, regular updates and well-informed staff all reduce the risk significantly. 

Should Small Businesses Be Worried About AI Cyber Attacks? 

Aware, yes. Worried, no. Every few months a fresh headline suggests AI has handed cybercriminals a superpower and small businesses are doomed, which is a great way to sell newspapers and a poor way to run a business. The reality is more measured. The UK’s National Cyber Security Centre does expect AI to make attacks more frequent and more effective over the next couple of years, but what it describes is existing attacks getting a polish, not a brand new species of threat arriving from nowhere. 

What Has AI Changed for Cyber Criminals? 

The biggest change is quality. For years, one of our best defences against phishing was that the emails were clumsy, full of odd phrasing and spelling mistakes. AI tools have largely removed that giveaway, so a criminal can now produce a polished email in perfect British English, tailored to your industry, in seconds. Voice cloning is the other development worth knowing about, because a short clip from a website video or voicemail greeting can be enough to copy someone’s voice, and there are real overseas cases of staff fooled by fake calls and even fake video meetings. To keep this in proportion, those cases mostly involve large companies and big sums, and official UK figures don’t yet record how often AI is involved. For a typical small business, the most likely scenario is still a good-looking fake email asking for a login or a payment. 

What Has Not Changed About Cyber Attacks? 

The reassuring part is that, however clever the message, the attacker still needs one of three things from you: a click on a bad link, a login typed into a fake page, or a payment sent to the wrong account. Sensible habits and good tools are designed to stop exactly those things, and they work just as well against an AI-written email as a human-written one. Roughly four in ten UK businesses report a cyber incident each year, with phishing leading the list, so the sensible response to AI is getting the basics properly in place rather than chasing every new headline. 

How Can a Small Business Reduce the Risk? 

Start with your logins. Multi-factor authentication means a stolen password isn’t enough on its own, and passkeys, which replace passwords with your fingerprint or face, go further because there is nothing to type into a fake page. Both are free on most platforms and take minutes to set up. Next, look at your email protection. Basic filtering struggles with new, well-crafted messages, whereas the advanced protection in plans such as Microsoft 365 Business Premium checks links at the moment of clicking and tests attachments before they reach an inbox, which is exactly the job you want done when an email looks flawless. 

If anyone receives a request to change bank details or pay an urgent invoice, it should be checked by phone, using a number you already hold rather than one in the email, before any money moves. For larger sums, a second person approving it is better still. That one habit defeats fake emails, cloned voices and hijacked supplier accounts alike, because it doesn’t rely on anyone spotting the fake. Make it clear that pausing to check is always welcome, even when the boss is the one apparently asking. 

Finally, keep devices and software updated, keep a backup stored outside your main systems, and give staff a simple rule for AI tools: nothing confidential, such as client details or financial figures, goes into a free chatbot. It’s easy to forget that the convenient assistant is also a third party. 

The Sensible Conclusion 

AI has mostly made criminals’ emails prettier rather than their methods cleverer, so the businesses that cope best will be the ones with the dull basics sorted. If you’d like to talk through where your business stands, or want a second opinion on how well protected you are, get in touch with the team at Lucid Computer Solutions and we’ll give you a straight, jargon-free answer. 

FREQUENTLY ASKED QUESTIONS 

Are small businesses really being targeted by AI cyber attacks? 

Yes, but mostly in an everyday way. Criminals use AI tools to write more convincing phishing emails and fake invoices and to send them to large numbers of businesses at once, so size doesn’t protect you. The attack methods themselves, such as phishing, fake payment requests and stolen logins, are the same ones small businesses have faced for years, which is why the standard protections still work. 

What is a deepfake and should I be concerned about them? 

A deepfake is a fake voice recording or video, created with AI, that imitates a real person. Most high-profile cases have involved large organisations, so for a small business the risk is real but fairly low. The simple defence works against them too: any unusual request for money or sensitive information should be checked by phoning the person on a number you already trust, however convincing the call or video seemed. 

Can AI-written phishing emails be spotted? 

It’s harder than it used to be, because the spelling mistakes and awkward wording have largely gone. It helps to focus on the request rather than the writing. Be cautious of anything urgent, unexpected, or asking for a payment, a password or a change of bank details, and check it through a separate channel before acting. 

Does multi-factor authentication still protect against AI attacks? 

Yes. Multi-factor authentication stops a stolen password from being enough to get into an account, which holds true however the password was stolen. Passkeys go a step further by removing the password altogether. Both are among the most effective and affordable protections available to a small business. 

Is it safe for staff to use AI tools like ChatGPT at work? 

It can be, with a sensible rule in place. Free AI tools may use what you type to improve their models, so staff should avoid entering client details, financial figures or anything confidential. Business versions of AI tools usually have clearer data protection terms, and it’s worth agreeing which tools are approved so the team isn’t guessing. 

Gavin Moorhouse is the CEO of Lucid Computer Solutions, a leading Worcestershire-based IT Services business. Give them a call on 01527 908646. 

Comments are closed.