Uncategorized

Last week I stumbled across a fascinating BBC article — “Cyber attack contingency plans should be put on paper, firms told” (14th Oct).Yes, you read that right. The government is apparently urging CEOs across the country to print their cyber contingency plans. On actual paper. I guess it’s news because it could be seen as a backwards step, it certainly feels like it should be but with cyberattacks on the rise (many now fuelled by AI), perhaps the pendulum is starting to swing back towards the analogue world. When “Offline” Was the Safety Net Back in the day, we provided DRIPs — Disaster Recovery Implementation Plans — complete with printed “break glass” copies. They lived in fireproof safes or desk drawers, ready for the worst: fire, theft, ransomware, you name it. Inside those spiral bound documents were all the essentials:Who to call (with phone numbers), the communication chain, how to reconnect to backups, how far back data could be recovered — it was all there. Of course, paper plans had one flaw — they were outdated almost before the A4 hit the Laserjet People changed roles, systems evolved, licenses shifted. So we moved to digital DRIPs, which made sense… until you realise what happens if you can’t access them. When the Backup Becomes the Target Imagine your digital recovery plan gets crypto-locked, or worse — stolen.Now the attacker knows exactly how you plan to recover, who’s critical, and what systems you can’t live without. That’s not just a breach — it’s giving your opponent the map, the key, and the alarm code. And yet, many businesses are doing just that as they sprint towards cloud-only infrastructures.Sure, cloud backups are convenient — but are they safer than an old-school, air-gapped tape drives?I still remember clients who “forgot” to swap tapes everyday or left them on the office desk (a classic), but at least those tapes couldn’t be encrypted from 5,000 miles away. Maybe the Old Ways Still Work Physical keys can’t be hacked.Printed QR codes can’t be spoofed. NASA knew this decades ago — that’s why the Apollo missions had manual overrides and physical backups for navigation. I bet the astronauts were very glad of that when things got bumpy. And it’s not just business continuity. Think about your own data.All those thousands of photos stored on your phone — will you still be able to open them in 10 years’ time?Printing a few key memories in a photo book might feel old-fashioned, but it’s also future-proof. (For the record, I’ve got a shelf full of them — highly recommended!) The Case for a Little Analogue I’m not saying we ditch digital. It’s faster, smarter, and more connected than ever. But maybe — just maybe — we should keep a physical copy of the things that really matter. Whether it’s your company’s cyber recovery plan or a picture of Tiddles the cat sitting proudly on little Susie’s pram in 1987, don’t assume the tech will be there when you need it most. Now, if you’ll excuse me… I’m off to print our boarding passes for the half-term trip to Majorca. Simon

AI in the Workplace: What’s Everyone Really Doing? One question I hear a lot when talking about technology change — especially AI — is: “So… what’s everyone else doing?” It’s a fair question. Sometimes it feels like everyone’s caught up in the AI hype without really knowing why. That’s not new — we’ve seen this “hype stage” before in the Gartner Hype Cycle, where excitement is followed by the inevitable trough of disillusionment. But AI feels different this time. Scroll through LinkedIn and you’ll see vendors and MSPs gushing about their brand-new “in-app” AI features delivering “mega value.” But strip away the marketing spin and ask:  Who’s actually using AI — and how are they benefiting? I’ve talked a lot about Microsoft Copilot, seen demos where an AI agent can help a shopper pick the right golf club based on their swing speed, or suggest the perfect candle scent for a dinner party vs. a study session. It’s fun, it’s clever even — but does it translate into measurable sales ? Some companies are now moving past the “let’s have a look” stage. They’re assigning budgets for Proof of Concept projects, and even creating AI Steering Committees. But that brings us back to the real question: Where are the benefits? When Copilot first launched, the focus was all about end-user perks: That all sounds great — but did it actually make or save the company money? If someone is 5% more efficient in their daily tasks, does that boost the bottom line? Or does it just mean they spend an extra 15 minutes a day scrolling through cat videos on TikTok on the toilet? (Boosting a different kind of “bottom line.”) Moving to the Back End: Agents & Studio The conversation has now shifted to back-end gains with features like Agents and Studio, where you can pre-prompt AI to handle repetitive tasks and queries. Imagine onboarding a new starter without the usual chaos: Or picture HR creating a SharePoint hub with policies on expenses, company cars, payroll, parking, visitor processes, health & safety, etc. An Agent can then answer natural language queries like: Does that save enough time to boost productivity? Possibly. Does it reduce HR interruptions and therefore free up HR resource? Definitely. The Shadow AI Problem Maybe the bigger question isn’t “How should we use AI?” but “Is AI already being used without us knowing?” The answer is almost certainly yes.This “shadow IT” trend isn’t new — remember when employees used WhatsApp for work before Teams Chat existed? Or Dropbox before OneDrive? People used these tools because they were useful, eveif they were (unintentionally) leaking company data. The fix wasn’t to ban them — it was to provide sanctioned alternatives that offered the same benefits without the security risks. Sanction It or They’ll DIY It Anyway AI is now part of everyday life. People will use it at work, whether you provide it or not. If you don’t offer a safe, compliant option, they’ll just bring their own. Even if you do roll out something like Copilot for Microsoft 365, expect that employees will also use ChatGPT for personal tasks — and that’s fine, as long as company data stays in Copilot, where it’s protected inside your tenant. Bottom line: The question isn’t whether AI will be used in your business — it’s how it will be used, and whether you’re giving people a secure, approved way to do it that leaves you in control and your data secure. Whether time saved by staff translates to more output, higher profitability or simply more cat videos watched on the loo depends just as much on how you integrate AI into your business culturally as it does technically, but that’s for another post… Now, where’s that cat on roller-skates video… Simon.