📝 Show notes
CyberAware Daily — 2026-09-20
- Weekly recap: DarkWatch stats for the week: 22,214 captures, 11,775 entities, 1,806 alerts, 11 active targets.
- Dark web leak site activity from DragonForce group noted; no personal action needed for most households.
- Researchers used Claude Opus 5 to find chained flaws in OpenAI staff accounts; reported and fixed.
- Upcoming webinar on proving CVE exploitability before attackers do.
- Next week: Patch Tuesday expected with critical Microsoft fixes.
- Actionable takeaway: check your email and banking passwords, use strong unique passwords and turn on two factor authentication.
More at https://cyberawareuk.co.uk
📄 Full transcript
It's Sunday, so we're doing our weekly recap.
I am Bully, that's Alex, and we're going to look back at the week in scams and security, all from real data we trust.
And we're starting with numbers from Darkwatch, which tracks what's actually happening on the dark web.
Right. This week, Darkwatch captured 22,214 records across 11,775 entities.
That's a lot of noise, but it boils down to 1,806 alerts and 11 active targets.
So what does that mean for a normal family?
Those 11 active targets are the ones being actively watched or attacked right now.
It's not about you personally unless you happen to be one of those organizations.
Exactly.
And one of the captures was a post on an active dark web leak site from a group calling themselves Dragonforce.
We won't read out the address, but it's a reminder that ransomware crews are still posting stolen data when they don't get paid.
And what should our listeners take from that?
If you run a small business, make sure your backups are separate and tested.
If you're just a household, the main thing is to keep your devices updated and be careful with links.
You're not the target here, but good habits protect you anyway.
Now, top news this week.
Researchers used a new AI model, Claude Opus 5, to find chained flaws that let them take over OpenAI staff accounts.
That sounds scary, but it's actually a good thing.
Right.
The researchers were testing defenses, and they reported the issue so they could be fixed.
It shows that AI can help find problems faster, but it also means attackers might use similar tools.
The lesson for us is simple, use strong, unique passwords and turn on two-factor authentication everywhere you can.
And there's a webinar coming up about proving whether a new CVE is exploitable before attackers do.
That's more for IT teams, but it highlights that patching quickly still matters.
Absolutely.
We often say it, but it bears repeating, when updates come out, install them.
It's the easiest win.
We also covered a lot this week.
A deep read on UK fraud networks.
A research paper on voice clones needing just three seconds of audio, and a couple of Microsoft fixes.
Plus, guides on fake parcel texts, and what to do in the first hour after a scam.
If you missed any of those, they're all on our website.
And next week, we're expecting Patch Tuesday, with critical fixes from Microsoft.
So keep an eye out for updates on Tuesday or Wednesday.
Before we go, let's do our one actionable takeaway for the week.
This week, it's about passwords.
Take five minutes today to check your most important accounts, email and banking, and make sure they have a strong, unique password and two-factor authentication turned on.
If you haven't already, do it now.
It's the single best thing you can do to protect yourself.
That's it for today.
Thanks for listening, and stay safe.
See you tomorrow.