VoIP Providers
At the last meeting, we discussed switching your landline to VoIP (Voice over Internet Protocol), with the old PSTN (Public Switched Telephone Network) analogue phone system being switched off at the end of January next year. Most customers who wish to retain a “landline” service take the VoIP digital phone service offered by their broadband provider. This may well be the best approach and is certainly the simplest, but there are a number of independent providers of VoIP phone services. In fact, VoIP services have been available for many years, although they have mainly been used by businesses – examples include Sipgate (founded in 2004) and Vonage (founded in 1998).
In effect, VoIP is just a service carried over your Internet in the same way as Zoom is a service. VoIP that is independent of your ISP means that subsequently changing your ISP will not affect your phone service. This article from Switch My Broadband discusses the pros and cons of using your ISP or an alternative provider to supply your digital phone service:-
https://www.switchmybroadband.com/separate-voip-home-phone/
This article mentions Voipfone and Andrews & Arnold as VoIP providers, but some others are listed in this article, “Best VoIP Service of 2025”, from TechRadar:-
https://www.techradar.com/best/best-voip-service-for-home
While I am not recommending a switch to an independent VoIP provider, I wanted to explain that it is possible.
Windows 11 Updates
AI is continuing to find more bugs in all software, including Windows 11. The Patch Tuesday update in October (13th) fixed a record 1,416 flaws – that’s more security bugs fixed in one month than in the whole of last year! At some point, as AI finds virtually all the legacy bugs, the number of patches will start to decline, but we have not reached that point yet. For more details on the October Patch Tuesday, see this PCWorld article:-
https://www.pcworld.com/article/3260668/microsoft-breaks-record-fixing-1416-security-flaws-update-windows-now.html
This month’s Patch Tuesday (yesterday, 10th Nov) fixed another 1,235 flaws, according to this report from Bleeping Computer:-
https://www.bleepingcomputer.com/news/microsoft/microsoft-november-2026-patch-tuesday-fixes-1-235-flaws-outlook-zero-day/
Still a lot, but not quite as many as last month, so maybe the AI bug-finding has reached a plateau?
26H2
The update to Windows 11 version 26H2 is also now available. You can check your current version by going to:-
Settings (Windows Key + I), System, About
Look under the Windows info heading.
If you are not yet at 26H2, head to:-
Settings, Windows Update
Here you may see the option to download and install 26H2.
Note that support for Windows 11 24H2 ended on 13th October this year, so it won’t get any more Patch Tuesday updates. If you don’t see the option to upgrade to 25H2 or 26H2 in Settings, Windows Update, you could use the Windows 11 Installation Assistant, which you will find here:-
https://www.microsoft.com/en-gb/software-download/windows11
I would recommend backing up all your data before using this tool to update.
Scam Link Checker
Malwarebytes has introduced a free scam link checker:-
https://www.malwarebytes.com/scam-check/link
The general advice is: don’t click on links in emails and messages.
However, sometimes it might be easier and quicker to click on a link – if you can be sure it is safe. This new tool from Malwarebytes allows you to copy the link and paste it into the search box on the above website. It will then analyse the link and report back on its safety.
The same site can also check a phone number – just select Check a number at the top of the page. This is probably less useful, as scammers can spoof different numbers each time they call, so the result cannot guarantee that the call was genuine.
You can also use VirusTotal to check a web link. We have mentioned this a couple of times before, most recently at the September 2023 Club - see the relevant blog:-
https://computerclub100.blogspot.com/2023/09/
Dave asked how to ensure that you are checking the actual web address, rather than just the potentially deceptive displayed text. On a PC, you can do this by right-clicking on the text and selecting Copy link address or Copy (Hyper)Link.
Alexa+
Amazon introduced its AI version of Alexa, known as Alexa+, in the UK earlier this year. It was made available for free under their Early Access scheme. You can find out more and sign up for early access here:-
https://www.aboutamazon.co.uk/news/devices/introducing-alexa-in-the-uk-the-next-generation-of-alexa
After the Early Access scheme ends, Alexa+ will cost £19.99 a month. However, it will be included in Prime membership, which includes Prime Video, free next-day delivery and more for just £8.99 a month, so it’s difficult to see why anyone would subscribe to the much more expensive stand-alone version.
AI Fake Detector
Google is hosting a service called SynthID Detector at:-
https://synthid.com/
You need to sign in to use the detector, but you can currently sign in with Google, Apple or ChatGPT credentials. Because you have to sign in, Google can limit your daily usage to around 10 images/videos/audio files.
SynthID is a watermarking technology from DeepMind, Google’s AI research lab, which can be embedded into images, videos, audio and text files. See this short (2-minute) YouTube video for more:-
https://www.youtube.com/watch?v=9btDaOcfIMY
The technology is currently used by Google, OpenAI, Nvidia and Kakao (a South Korean conglomerate). You can upload a file to the SynthID website and it will detect the hidden watermark if it is AI-generated. Apple will also start to use this technology soon and other AI companies may follow.
The above website cannot detect AI-generated images, videos and audio produced by AI products that do not use the SynthID technology, so its usefulness is somewhat limited at the moment. See this Ars Technica article for more:-
https://arstechnica.com/ai/2026/10/google-rolls-out-improved-synthid-ai-content-detector-now-available-globally/
Jev – A Different Type of AI
Developed by a startup company called TypeSafe AI, Jev is a new class of AI (which they call a System One Model) that can only make decisions and select choices. Unlike the other AI systems we have become (somewhat) familiar with (such as ChatGPT, Gemini and Claude), Jev is not based on a Large Language Model (LLM). It does not generate text or images but instead makes choices in essentially three ways:-
1. It can make yes/no decisions.
2. It can select from a choice of options.
3. It can score or grade against a scale (i.e. a confidence rating).
It can do these things much faster than an LLM-based AI would.
Jev was only launched on 15th September this year, but has already been valued at $7.5 billion, with TypeSafe AI claiming that a third of Fortune 500 companies are already using it.
This recent article by Zapier (an American software company) explains a lot more about it:-
https://zapier.com/blog/jev/
Also, this 15-minute YouTube video from IBM offers an overview of how Jev works in comparison to LLMs in relatively simple terms:-
https://www.youtube.com/watch?v=YGgNBcIgI4s
Of course, it doesn’t take long for other companies to come up with competitive products. According to this TechCrunch article on 22nd October, OpenAI has released its own Decisions API model:-
https://techcrunch.com/2026/10/22/openai-makes-its-jev-competitor-widely-available-as-decision-model-competition-heats-up/
These decision models, like Jev, will probably work alongside LLM-based AIs to create a quicker workflow.
PHP Flaw
You will have heard on the news that a flaw in the scripting language PHP was discovered and leaked before everyone was ready. PHP is used on around 70% of all websites, and this vulnerability could allow hackers to access affected web servers. The bugs in PHP that enable it to be hacked have been in the software since 2012! Some sites had to be taken down while updates were applied to fix the problem.
The vulnerabilities were found by researchers at Wiz Research and Calif using AI tools in collaboration with OpenAI and Google. A plan was underway to allow time for sites to update their software, but the details were (accidentally) leaked a couple of days early, causing a global panic to update or protect thousands of web servers.
The faults were very serious, allowing an attacker to break into any affected web server without needing any authentication (e.g. a password).
This TechCrunch article has more details:-
https://techcrunch.com/2026/10/24/php-ocalypse-how-two-critical-web-bugs-almost-broke-the-internet/
Linux Kernel Flaw
Researchers at Anthropic, using their AI tools, found a critical vulnerability in the Linux operating system, which is used by over 60% of websites. In theory, the flaw allows a remote hacker to perform commands as ‘root’ (i.e. the highest level of privilege), which sounds very serious. However, in practice, it would be extremely difficult to implement such an attack. Patches are being issued by affected Linux distributions to address the issue.
See this Bleeping Computer article for more detail:-
https://www.bleepingcomputer.com/news/security/critical-linux-kernel-flaw-allows-remote-root-attacks-on-mptcp-servers/
Binance Crypto Theft
Binance, a large cryptocurrency exchange platform, was hit by the largest ever crypto theft of around $6.3 billion. This seems to be an ‘old-fashioned’ phishing attack which compromised one of their engineer’s laptops, rather than an AI-based attack. The theft is believed to have been carried out by a North Korean hacking group. See this Malwarebytes article for more:-
https://www.malwarebytes.com/blog/news/2026/10/binance-hit-by-largest-ever-crypto-theft-with-6-3-billion-stolen
Verify Amazon Messages
Amazon have introduced a verification system to check whether a message you have received, which appears to be from Amazon, is genuine. Simply forward a suspicious email or submit your question to:-
verify@amazon.co.uk
To learn more about Amazon’s scam prevention and protection, go to:-
amazon.co.uk/scams
Next Session
Wed 9th December 2026 at 2pm by Zoom
