AI in Today’s IT Landscape

And Why People Still Pose The Biggest Risk

By Lewis Hand, IT Pre-Sales Executive

December, 2025

Artificial intelligence has exploded over the last few years. Back in early 2023, ChatGPT hitting 100 million users felt insane. It reached this milestone in a shorter timeframe than TikTok, Instagram and WhatsApp. Fast forward to 2025 and it sees roughly 700 million people every week, with predictions putting that number close to a billion by the end of the year. And that’s just one platform. Add Copilot, Google Gemini and all the other options, and you start to see how deeply AI sits in everyday work now. If you need an example of how quickly AI is improving, see the below AI generated image of “Will Smith eating spaghetti” from just two years ago.

It’s not very convincing. Compare this to today, where often, people aren’t even sure if what they’re looking at is real or not.

But AI isn’t just there for asking random questions or generating spaghetti consuming imagery…

It’s running automation, crunching data, supporting cybersecurity tools and generally does a lot of jobs faster than we ever could. Despite all of that, one thing hasn’t changed: people are still the easiest way in for attackers. I see this constantly with customers. Most incidents don’t come from some Hollywood-style hack – it’s someone opening a very convincing fake email or clicking a link that looks genuine. With AI now being used by criminals, these emails and websites are miles more believable than they used to be.

Why human error still leads the charts

Roughly 95% of breaches come down to a person making a mistake. Weak passwords, responding too fast to an email that feels urgent, skipping MFA – it’s the same pattern across businesses of every size. The latest UK Cybersecurity Breaches Survey backs this up: phishing is still the number one cause of trouble. I regularly meet businesses who have lost thousands because of a single email that caught someone on the wrong day. One wrong click can undo the best security setup in the world.

The Best Approach: Solid Tech Backed By Trained People

Use The Right Tools

Endpoint protection and email filtering
AI-driven threat detection that spots unusual behaviour (built-in to some mail filtering platforms)
Multi-factor authentication everywhere it matters

Train Your Team Properly

Regular phishing campaigns to test your staff
Reminders to slow down before clicking anything unfamiliar
Cyber awareness shouldn’t be something you check off once a year. It needs to be part of people’s normal working habits nowadays

Keep Reviewing

Reviews turn incidents into improvements instead of repeats
Regular checks catch small gaps before they become disasters
Attackers don’t sit still, so neither can your systems or your people

Welcomm Security That Works In The Real World

We give customers both sides of the defense: the tech and the user training. Alongside email security and endpoint protection, we run awareness campaigns that mimic real attacks, so staff get used to spotting the red flags before the damage is done.

If it’s been a while since your last security review, now’s the time. Speak to our IT team or your account manager and we’ll help you tighten your setup and your staff habits. Even in an AI-driven world, people still make the biggest difference.