AI policy people are arguing again about how open powerful models should be, after an Anthropic report flagged the open model GLM-5.3 as a major cyberattack risk. For a small IT team, this is a reminder that open-source AI tools may be great for cost and control, but you should treat anything security related as …
Read more “AI risks, dead startup data, and over-trusting refusals”
Oaklands College is using virtual reality so plumbing students can revisit practical workshop demonstrations that their tutors have recorded in advance, giving them a way to replay skills teaching on demand outside the lab. For IT, this is a live example of VR content that needs storage, network bandwidth, and device management that will sit …
Read more “VR plumbing, youth hubs and estates planning in FE”
AI safety rules are under the spotlight after Microsoft vice chair Brad Smith warned that your washing machine is subject to more safety rules than many AI systems, and called for more AI kill switches along with clearer regulation, saying the UK is well positioned to lead on this. If you are testing or rolling …
Read more “AI kill switches, Beyond Zero security and UK ransomware spikes”
Attackers moved on a new SonicWall Secure Mobile Access 1000 flaw within days of disclosure, with at least 135 exploitation attempts seen against CVE-2026-102255, a CVSS 10.0 pre-auth SSRF bug that lets a remote attacker make the appliance issue requests on their behalf. If you run SonicWall SMA 1000, you need to verify patches are …
Read more “SonicWall exploits, tainted Androids and WordPress hits”
FE teachers are starting to face the learner challenge of “if the software is intelligent, why do I need to learn this?”, as set out in a piece on AI and assessment in further education by Vanessa Aradia. IT teams should expect sharper questions about why certain tools are blocked or allowed, so it is …
Read more “AI questions, pressure planning and skills gaps in FE”
AI agents are starting to act more like independent staff, and Computer Weekly reports that process mining is being used to track what they actually do and stop automation from going rogue. For a small IT team, this is a reminder that if you are trialling AI agents, you need clear logs, approvals, and a …
Read more “AI agents, NetScaler patches and a mini PC worth a look”
Apple is expected to unveil a new HomePod mini 2, an updated Apple TV 4K, a new HomeView hub, and smart home accessories with LG at its upcoming home event. If your offices use Apple gear for meeting rooms or reception, you may want to hold off any speaker or TV box refresh until we …
Read more “Apple home and iPad updates that might hit your budget”
AI policy in health is moving again, with the government accepting the MHRA’s recommendations on the use of AI in healthcare, after patients said they wanted AI that actually improves care, as reported by UKAuthority. For school IT, this is a sign that regulators are starting to pin down expectations around AI risk, data, and …
Read more “AI pressure points for UK education IT this week”
Typosquatters are abusing rare Cyrillic and Latin letters in Chromium browsers to spoof popular domains, as shown in a recent analysis of how two characters can open up many fake lookalike sites. This means user training on URL checking is less useful on its own, so you should rely more on managed bookmarks, email link …
Apple has added the original 2017 Apple TV 4K to its vintage products list. If you still have these in meeting rooms or reception areas, plan for limited future support and start lining up a replacement cycle. macOS 27 brings remote AppleCare log collection as a behind the scenes upgrade. This matters for small IT …
Read more “Apple TV 4K goes vintage as macOS 27 boosts remote support”