Wednesday, 4 November 2026
Wednesday, 4 November 2026
- In brief
- 1Hackers warm supermarket fridges in at least 11 European countries and reach British meat suppliers, and the EU plans a joint response
- 2Binance says a fake job interview began its theft
- 3A ballot scanner fault feeds voting rumours
- 4Man charged over 700 powder letters to officials
- 5Police shut a firm that sold fake trading sites to scammers
- 6Password manager mSecure tells users to change everything
- 7AI attack tool reaches two more firms
- 8Red Hat fixes a top-rated Linux flaw; Ubuntu still testing
- 9Bootkit beats fully updated Windows computers
- 10German lawmakers among Apple spyware targets
- 11Harvard breach counted at 851,000 email addresses
- 12School software breach reaches 1.3 million people
- 13Hackers push a backdoor through a security tool's updates
- 14Barracuda patches two flaws that give full control
- 15Ransomware hits UK schools in a suspected AI-run attack
- 16Russia detains 3 in a rare hacking arrest
- 17Defence firm pays $13 million over cyber rules
- 18CrowdStrike buys two firms that check AI agents
- 19Android update fixes over 200 flaws
- 20NATO drill puts AI agents into cyber training
- 21Russia-aligned spies target Ukrainian courts and councils
- 22Fake bank sites pretend to be shut down
- 23US agency confirms attacks on a WordPress flaw
- 24Security investor Ballistic seeks $490 million
4 Nov · Cyber · In brief
Cyber · Wednesday, 4 November 2026
Cyber, in brief
24 stories · 43 sources
Swipe up for the day in brief ↑Scroll down for the day in brief ↓
What is moving · 1 of 4
Old fixes not installed
British meat suppliers were hit through fridge controllers whose fix had been out since August.
What is moving · 2 of 4
Attackers using AI tools
WorkOS says an attacker used ARTEX, an AI attack tool, to reach its sales app.
What is moving · 3 of 4
People left guessing
Password app mSecure told its users to change their passwords without saying what happened.
What is moving · 4 of 4
Police catching suspects
An operation involving police from seven countries arrested 7 people behind fake trading websites.
24 stories
- 1Hackers warm supermarket fridges in at least 11 European countries and reach British meat suppliers, and the EU plans a joint response
- 2Binance says a fake job interview began its theft
- 3A ballot scanner fault feeds voting rumours
- 4Man charged over 700 powder letters to officials
- 5Police shut a firm that sold fake trading sites to scammers
13 sources
Cyber, in brief
- 1EU urges action as Germany, Romania and UK report supermarket cyberattacks
- 4British businesses caught in Europe-wide supermarket fridge attacks
- 7Supermarket Chain Thaw Massacre: Russia suspected in Europe's fridge attacks
- 10Binance CEO says 'months-long campaign' led to $6.3 billion theft as N. Korea denies responsibility
- 14The Internet's Conspiracy Theorists Are Desperate for Voter Fraud
- 16FBI arrests man for sending over 700 threatening letters to US officials
- 18Germany Hits 'Crime-as-a-Service' Investment Scam
- 20Password manager mSecure tells users to change passwords, won't say why
- 22WorkOS says attacker used ARTEX AI to breach internal sales app
- 24Red Hat Patches 9.8-Rated Linux TCP RCE Flaw While Ubuntu Releases Fixes for Testing
- 25UEFI bootkit infected fully patched Windows machines. Is Microsoft to blame?
- 32UK school estate hit by 3AM ransomware in suspected AI-assisted attack
- 33Russia dismantles infrastructure of VasyGrek cybercrime group, detains three suspects
Hackers warm supermarket fridges in at least 11 European countries and reach British meat suppliers, and the EU plans a joint response
Swipe up for the story ↑Scroll down for the story ↓
Hackers warmed supermarket fridges and freezers in at least 11 European countries on Monday, officials say.
Monday
nine EU countries report hacked fridges
Tuesday
Germany, Romania and Britain report theirs
Germany, Romania and Britain reported their cases on Tuesday, a day after nine EU countries did.
“These are not isolated incidents”
"These are not isolated incidents," said Thomas Regnier, a spokesperson for the European Union.
Who is involved
Danfoss and Copeland
Danish and US makers of shop fridge controllers; older or unpatched controllers of theirs were hit [4][5]
Who is involved
The NCSC
Britain's National Cyber Security Centre; it warned exposed firms overnight [4]
Why it matters
Shops threw away food, and some said the losses reached hundreds of thousands of euros.
Why it matters
Grocers keep only about 1 to 3 percent of their takings as profit, Scientific American notes.
Why it matters
Claroty, which found the controller flaws in August, warns attackers could next hit warehouses that supply many shops.
How it works
Refrigeration engineers call it a defrost cycle: heaters warm the cooling coils every so often, so ice cannot block them.
How it works
The attackers kept that heater switched on, so it warmed the whole freezer, in some shops to 50C.
Seen it before
A toaster told to stay on burns the bread, even though toasting is its normal job.
The truth
Lupovis
a Russian state-backed group
Romania's cyber agency
semi-amateurish, no state
Lupovis's chief executive says the attacks could be the work of a Russian state-backed group.
The truth
Belgium's cyber agency says it may never be possible to identify who did it.
Good news
A fruit and vegetable wholesaler, Langridge Organic, unplugged a fridge from the internet while it was being switched to defrost.
What comes next
The meat suppliers' group is asking Britain's government to treat food as critical national infrastructure.
How it works
Why a rumour grows when something matters and nobody explains it
How it works
Social psychologists say a rumour spreads with how much an event matters and how unclear its cause is, so a plain, early explanation slows it.
How it works · step 1 of 5
Something that matters to many people happens
How it works · step 2 of 5
Its cause is unclear, or the explanation keeps changing
How it works · step 3 of 5
People fill the gap with their own guesses
How it works · step 4 of 5
The guesses travel faster than any check
How it works · step 5 of 5
A clear, early account shrinks the gap
The twist
A rumour does not need a liar to start.
The twist
An event that matters plus a missing explanation is enough, and the explanation is the one part the people in charge can supply.
Seen the same day
The county gave one cause, withdrew it 20 minutes later, and posts calling it cheating reached 100 a minute.
Seen the same day
With no claim and no explanation of how attackers got in, experts' guesses ran from a Russian state group to semi-amateurs.
Seen the same day
The maker said something happened but not what, so one customer asked whether all 800 of their passwords had to change.
Seen the same day
The opposite move: a step-by-step timeline from the fake interview to the signing service left fewer gaps to guess into.
The catch
An explanation given too fast can be wrong, and a corrected one can feed the rumour it was meant to stop, as Westchester's changing statement did.
The whole
Each person who shared a Westchester post saw one jammed scanner and one guess, not the whole county's count.
The whole
Anyone reading an alarming headline before the cause is known is standing in the same gap.
What's really going on
The fridge attacks and the BlackX bootkit both went through holes that already had a fix.
What's really going on
The British meat suppliers' controllers had an update waiting since August, and Microsoft's Windows start-up fix works only after a manual step few administrators took.
What's really going on
The attackers needed no new flaw, only owners who had not finished the old work.
Why it works on people
A freezer heated to 50C makes an attack look new and clever, even when it went through a published hole with a fix already out.
Who pays
Shop managers and staff in the hit supermarkets
They came in early to bin the stock, and one Delhaize manager said it made him feel like quitting his job.
Nobody knows yet
Who heated Europe's shop fridges, and why.
No one has claimed it; one security firm's chief points to a Russian state-backed group, Romania's cyber chief calls it semi-amateurish, and Belgium's agency says it may never be known.
Nobody knows yet
How the attackers got into the fridge controllers on Monday.
Officials have not explained it publicly, and Censys did not say whether any patched devices were among those hit.
Nobody knows yet
Whether the freezer failures at US military bases in August were the same attackers.
The US Defense Department suspects hackers but has not said who, and investigators have not linked the two waves in Europe either.
Good news
Sweden's Coop, whose seven hit stores ran their fridges over old 4G links, has sped up moving its fridge equipment onto modern 5G modems.
The short version
That's the story in short.
The whole story has 7 more parts.
or swipe up for the next storyor scroll on for the next story ↓
The whole story, in 7 parts ↓
Part 1 of 7 · Who else is involved
The European Union
the union of European countries; its spokesperson says EU agencies are planning a joint response [1]
Part 1 of 7 · Who else is involved
Dan Cimpean
head of Romania's cyber security directorate; he called the attacks semi-amateurish [1]
Part 2 of 7
11-12 Oct
about 700 shops hit, mostly in France
Monday
the second wave, in at least 11 countries
This is the second wave: the first hit about 700 supermarkets, mostly in France, on 11 and 12 October.
Part 3 of 7
Belgium's federal prosecutors
opened a criminal investigation
Belgium was among the hardest hit, and its federal prosecutors have opened a criminal investigation.
Part 3 of 7
In Italy, about 20 supermarkets were hit, including four Co-op stores in Sicily.
Part 3 of 7
Germany had 4 supermarkets hit and Romania 6, officials and companies said on Tuesday.
Part 3 of 7
Italy's government said the shops hit mostly had old systems, open to the internet, with weak password checks.
Part 4 of 7
In Britain, three members of the Association of Independent Meat Suppliers were hit, the trade group says.
Part 4 of 7
“collateral”
Its spokesman Tony Goodger called them "collateral", because they ran the same fridge software as the shops.
Part 4 of 7
The fridges hit were empty at the time, so no meat was lost, the group told the BBC.
Part 4 of 7
NCSC
Britain's National Cyber Security Centre
Britain's cyber agency, the NCSC, worked through the night to warn firms whose systems sat open online.
Part 5 of 7
Copeland
US maker of fridge controllers
The British firms' fridges were run by XWEB controllers from Copeland, a US maker of refrigeration controls.
Part 5 of 7
XWEB firmware 1.12.1 and older let attackers "take control without a password", a note to the meat group's members says.
Part 5 of 7
August
Copeland's fix is out
Monday
firms on old firmware are hit
A fix had been out since August, but the firms hit were reportedly still running the old firmware.
Part 5 of 7
Danfoss
Danish maker of fridge controls
Danfoss, a Danish maker of fridge controls, told customers on 13 October to take their systems off the internet.
Part 5 of 7
Most stores
went offline
A small number
did not
Most stores did, but a small number did not, a Danfoss spokesperson told the Dutch broadcaster NOS.
Part 5 of 7
Researcher Kevin Beaumont says every freezer hit on Monday had a control page reachable from the internet.
Part 6 of 7
20 Oct
decoys see the first login attempts
26 Oct
attempts peak
Monday
the second wave
Lupovis, a security firm that runs decoy fridge controllers, saw logins tried with Danfoss's factory passwords from 20 October.
Part 6 of 7
The attempts peaked on 26 October, and the decoys refused every one of them.
Part 6 of 7
Exposed Danfoss control pages in France fell from 1,825 before the October attacks to 220, says Censys, an internet-scanning firm.
Part 7 of 7
No one has claimed the attacks.
Part 7 of 7
August
freezer failures at US bases
Oct and Nov
two waves in Europe
The US Defense Department now suspects hackers caused freezer failures at US military bases in August, Scientific American reports.
8 sources
Hackers warm supermarket fridges in at least 11 European countries and reach British meat suppliers, and the EU plans a joint response
- 1EU urges action as Germany, Romania and UK report supermarket cyberattacks
- 2Europe on alert after supermarket freezers hit by cyber-attacks in 10 countries
- 3UK meat industry hit by cyber-attack which targeted European supermarket freezers
- 4British businesses caught in Europe-wide supermarket fridge attacks
- 5Over 100 European Stores Hit as Supermarket Fridge Attacks Return
- 7Supermarket Chain Thaw Massacre: Russia suspected in Europe's fridge attacks
- 8How hackers are turning grocery store freezers into heaters
- 25UEFI bootkit infected fully patched Windows machines. Is Microsoft to blame?
The full story
Hackers warm supermarket fridges in at least 11 European countries and reach British meat suppliers, and the EU plans a joint response
Hackers warmed supermarket fridges and freezers in at least 11 European countries on Monday, officials say.
This is the second wave: the first hit about 700 supermarkets, mostly in France, on 11 and 12 October.
Belgium was among the hardest hit, and its federal prosecutors have opened a criminal investigation.
In Britain, three members of the Association of Independent Meat Suppliers were hit, the trade group says.
The British firms' fridges were run by XWEB controllers from Copeland, a US maker of refrigeration controls.
Lupovis, a security firm that runs decoy fridge controllers, saw logins tried with Danfoss's factory passwords from 20 October.
No one has claimed the attacks.
The meat suppliers' group is asking Britain's government to treat food as critical national infrastructure.
Why it matters. Shops threw away food, and some said the losses reached hundreds of thousands of euros.
Binance says a fake job interview began its theft
Swipe up for the story ↑Scroll down for the story ↓
Binance, a crypto exchange, published on Tuesday its account of how hackers took about $6.34 billion.
Software engineer
contacted in September
Fake recruiter
on a real job site
It says a software engineer was contacted in September by a fake recruiter, through a real job-hunting site.
In a second interview he was told to download a coding test, which hid new malware.
Antivirus tools did not spot the malware, Binance says.
By 20 October the attackers had used his computer and an internal AI tool to reach 2 more staff.
September
fake interview
20 Oct
2 more staff reached
29 Oct
the coins are taken
On 29 October they used stolen logins to send coins out through an internal wallet signing service.
How it got here
FBI
blames North Korea
North Korea
denies it
The FBI blames TraderTraitor, a North Korean state hacking group.
Why it matters
Binance says it stopped 57 other North Korean attempts this year, including fake job offers.
Why it matters
About half the coins were frozen, but billions are still held by the thieves, Wired reports.
How it works
Security teams call this social engineering: tricking a person, not a machine, as the fake recruiters did.
How it works
The fake team built trust over two interviews, with a video call and photos, before the download.
Seen it before
A stranger who chats kindly for weeks before asking to borrow your house key is using the same trick.
How it works
The opposite move: a step-by-step timeline from the fake interview to the signing service left fewer gaps to guess into.
The idea, in story 1: Why a rumour grows when something matters and nobody explains it →Who gains
TraderTraitor, North Korea's state hacking group
Billions in stolen coins that were not frozen are still in its hands while it launders them, Wired reports.
Who pays
Binance
It says it has set money aside to cover users' losses and rebuilt or replaced the systems hit.
3 sources
Binance says a fake job interview began its theft
The full story
Binance says a fake job interview began its theft
Binance, a crypto exchange, published on Tuesday its account of how hackers took about $6.34 billion.
Why it matters. Binance says it stopped 57 other North Korean attempts this year, including fake job offers.
A ballot scanner fault feeds voting rumours
Swipe up for the story ↑Scroll down for the story ↓
Ballot scanners across Westchester County, New York, would not accept paper ballots when polls opened on Tuesday.
Voters put their ballots into bags kept safe, to be counted later instead.
What comes next
Tuesday
ballots go into bags
Wednesday
the count is expected
A judge let officials move thousands of these ballots to a secure warehouse.
Posts on X about the fault reached as many as 100 a minute, a University of Washington researcher found.
Some posts claimed Democrats were cheating, and AP's fact check found no credible evidence of that.
“There's no shenanigans going on here”
Mike Lawler, the Republican whose seat was at stake, said: "There's no shenanigans going on here."
In Fulton County, Georgia, some voters got false texts about polling-place changes that the county never sent.
The truth
Westchester board
a paper issue
AP
a programming glitch
The county board first blamed thick ballot paper, then dropped that 20 minutes later for a general paper issue.
Why it matters
Mike Lawler
Republican, seeking re-election
Cait Conley
Democrat
The fault hit the hotly contested House race between Lawler and Democrat Cait Conley.
How it works
The county gave one cause, withdrew it 20 minutes later, and posts calling it cheating reached 100 a minute.
The idea, in story 1: Why a rumour grows when something matters and nobody explains it →Who pays
Cait Conley's and Mike Lawler's voters in Westchester
Their ballots wait in bags for a count expected to run until at least Wednesday.
Nobody knows yet
What made Westchester's ballot scanners fail.
The county first blamed thick paper and withdrew that within 20 minutes, while AP reports a programming glitch.
The full story
A ballot scanner fault feeds voting rumours
Ballot scanners across Westchester County, New York, would not accept paper ballots when polls opened on Tuesday.
Why it matters. The fault hit the hotly contested House race between Lawler and Democrat Cait Conley.
Man charged over 700 powder letters to officials
Swipe up for the story ↑Scroll down for the story ↓
Robin Patrick Martin
67, of Oklahoma City; charged
Robin Patrick Martin, 67, of Oklahoma City, is charged with mailing threatening letters with white powder to officials.
Prosecutors say he sent more than 700 letters to state and federal officials, including members of the US Congress.
Early tests found the powder was harmless yeast or flour, though not every letter has been tested.
Some letters carried messages in ink that shows only under ultraviolet light.
How it got here
Sep 2024
first letters arrive
Now
Martin is charged
The letters began reaching state election officials on 16 September 2024, headed "DECLARATION OF WAR".
Why it matters
Officials in nearly every US state received similar mailings over two years, CyberScoop reports.
How it works
Detectives call it following the paper trail: phone records led to a Walmart, where video showed a man matching Martin.
How it works
Records from the phone also placed Martin at postal centres around Oklahoma, investigators say.
Seen it before
A shop receipt in your pocket shows where you were and what you bought.
The truth
Jan 2025
Trump administration takes over the case
This week
Martin is charged
FBI Director Kash Patel said the Biden administration "failed to catch" Martin.
Good news
Federal agents arrested a 67-year-old Oklahoma City man accused of mailing more than 700 powder letters to officials, and the powder tested so far was yeast or flour.
2 sources
Man charged over 700 powder letters to officials
The full story
Man charged over 700 powder letters to officials
Robin Patrick Martin, 67, of Oklahoma City, is charged with mailing threatening letters with white powder to officials.
Why it matters. Officials in nearly every US state received similar mailings over two years, CyberScoop reports.
Police shut a firm that sold fake trading sites to scammers
Swipe up for the story ↑Scroll down for the story ↓
Europol, the EU's police agency, says police arrested 7 people and seized more than 100 servers.
The company
sold sites and call centres
Criminal gangs
bought them
The target was a company accused of selling fake trading websites and call centres to criminal gangs.
Cyprus and Israel
where the firm was run
Albania, Kosovo, North Macedonia
the call centres
It was run from Cyprus and Israel, with call centres in Albania, Kosovo and North Macedonia, Europol says.
How it got here
The case began in 2022, when a German couple could not withdraw money from an online trading platform.
Losses are put at more than 4 million euros, with possible victims in more than 50 countries.
Why it matters
Victims saw fake profits that pushed them into bigger transfers, then were cut off when they asked for money.
How it works
One firm
supplies the tools
Many gangs
rent them
Police call it crime as a service: one firm supplies the tools, and many gangs rent them.
How it works
Here the firm built the sites and sold stolen contact details, and the gangs' call centres did the phoning.
Seen it before
A shop that rents disguises to every robber in town is easier to close than each robber.
The truth
Dark Reading reports searches at 17 locations.
Good news
An operation involving police from seven countries arrested seven people and seized more than 100 servers used to sell fake trading websites to scam gangs.
3 sources
Police shut a firm that sold fake trading sites to scammers
The full story
Police shut a firm that sold fake trading sites to scammers
Europol, the EU's police agency, says police arrested 7 people and seized more than 100 servers.
Why it matters. Victims saw fake profits that pushed them into bigger transfers, then were cut off when they asked for money.
Password manager mSecure tells users to change everything
Swipe up for the story ↑Scroll down for the story ↓
mSecure, an app that stores people's passwords, emailed customers on 3 November to change their passwords.
It also suggested resetting two-step login checks and removing passkeys.
“suspicious activity”
Its maker, mSeven Software, said it found "suspicious activity" but did not say what happened.
It has not said when this happened, or whether anyone's stored data was reached.
Why it matters
One customer asked whether the email meant changing all of their more than 800 passwords.
How it works
Cryptographers call this end-to-end encryption: mSecure says vault data is scrambled on your device before syncing.
Seen it before
A diary whose lock only you have the key to stays private, even from the shop that sold it.
The truth
Germany's BSI found in 2025 that one older Android version could, in theory, let its maker read stored data.
How it works
The maker said something happened but not what, so one customer asked whether all 800 of their passwords had to change.
The idea, in story 1: Why a rumour grows when something matters and nobody explains it →Who pays
mSecure's customers
Without knowing what happened, some face changing hundreds of saved passwords by hand.
Nobody knows yet
What actually happened at mSecure.
Its maker has not said what the suspicious activity was, when it happened, or whether stored vaults were reached.
2 sources
Password manager mSecure tells users to change everything
The full story
Password manager mSecure tells users to change everything
mSecure, an app that stores people's passwords, emailed customers on 3 November to change their passwords.
Why it matters. One customer asked whether the email meant changing all of their more than 800 passwords.
AI attack tool reaches two more firms
Swipe up for the story ↑Scroll down for the story ↓
WorkOS
runs sign-in tools for OpenAI, Anthropic and others
WorkOS, which runs sign-in tools for apps such as OpenAI and Anthropic, says an attacker reached an internal sales app.
The attacker took business contacts and summaries of customer calls and chats, but no app data, WorkOS says.
WorkOS says the attacker used ARTEX, an AI tool that automates the steps of an attack.
Chainstack
runs servers for blockchain apps
Chainstack, which runs servers for blockchain apps, says an attacker reached parts of its live systems through a third-party component.
Customer names, emails and password hashes, scrambled forms of passwords, were exposed, Chainstack says.
How it got here
Chainstack's break-in overlaps with a China-linked campaign using ARTEX, The Hacker News reports.
26 Oct
attacker gets in
31 Oct
access cut off
The attacker got into a development system on 26 October, and Chainstack cut it off on 31 October.
Why it matters
Chainstack asks customers to change node keys, API keys and any reused passwords.
How it works
Security staff call this automated reconnaissance: software scanning a target for weak spots, as WorkOS says ARTEX did.
Seen it before
A burglar with a robot that tries every door on a street at once needs far less time.
The truth
Chainstack says it has no evidence the node keys were read, but treats them as exposed.
2 sources
AI attack tool reaches two more firms
The full story
AI attack tool reaches two more firms
WorkOS, which runs sign-in tools for apps such as OpenAI and Anthropic, says an attacker reached an internal sales app.
Why it matters. Chainstack asks customers to change node keys, API keys and any reused passwords.
Red Hat fixes a top-rated Linux flaw; Ubuntu still testing
Swipe up for the story ↑Scroll down for the story ↓
- 1The flaw's rating9.8 out of 10
Red Hat, which sells a business version of Linux, shipped fixes on 1 November for a flaw rated 9.8.
An attacker can trigger it by sending connection requests to any open port, with no login needed.
Red Hat first rated it 5.9, then raised it to 9.8 and called it critical.
Ubuntu's fixes are still only in an opt-in testing channel, with no release date.
How it got here
Linux 4.4
the bug arrives
October
the main fix lands
1 Nov
Red Hat ships fixes
The bug has been in Linux since version 4.4, and the main fix landed in October.
Why it matters
Servers need a restart after the update, so owners must plan for downtime.
How it works
Programmers call it a use-after-free bug: the system reads memory it has already handed back.
How it works
Here a repeated connection request arrives while another processor still reads options already freed.
Seen it before
Reading a note you already threw away, after someone else wrote on it, gives you the wrong message.
Good news
Fixed versions are out for Red Hat's 3 current releases, 8, 9 and 10.
1 source
Red Hat fixes a top-rated Linux flaw; Ubuntu still testing
The full story
Red Hat fixes a top-rated Linux flaw; Ubuntu still testing
Red Hat, which sells a business version of Linux, shipped fixes on 1 November for a flaw rated 9.8.
Why it matters. Servers need a restart after the update, so owners must plan for downtime.
Bootkit beats fully updated Windows computers
Swipe up for the story ↑Scroll down for the story ↓
Kaspersky
security firm; found BlackX
Kaspersky, a security firm, found a bootkit called BlackX on Windows computers with every update installed.
A bootkit is malware that starts before Windows does, when many defences are not yet running.
It uses Baton Drop, a flaw in Windows start-up that Microsoft fixed in 2022.
How it got here
2022
Microsoft fixes the flaw
2023
BlackLotus uses it
Now
BlackX uses it
BlackLotus, a bootkit found in 2023, used the same flaw.
Microsoft's 2024 fix blocks old start-up files only if administrators take an extra manual step.
Why it matters
Kaspersky says few organisations knew the step mattered, and the attackers counted on that.
How it works
Engineers call this revoking a certificate: telling a computer to stop trusting an old signed file.
How it works
Microsoft did not force it, because revoking too early could leave millions of PCs unable to start.
Seen it before
A school that changes its locks but lets the old keys keep working has not really changed them.
The truth
Microsoft's guidance from two weeks ago does not warn that the patches alone leave this gap, Ars Technica reports.
1 source
Bootkit beats fully updated Windows computers
The full story
Bootkit beats fully updated Windows computers
Kaspersky, a security firm, found a bootkit called BlackX on Windows computers with every update installed.
Why it matters. Kaspersky says few organisations knew the step mattered, and the attackers counted on that.
German lawmakers among Apple spyware targets
Swipe up for the story ↑Scroll down for the story ↓
German lawmakers and staff
warned by Apple
German lawmakers and foreign-policy staff were among people Apple warned last week about spyware, Germany's BSI says.
BSI
Germany's information security agency
The BSI is Germany's information security agency.
Green Party co-leader Franziska Brantner said on Tuesday that she was one of them.
Many were travelling with Chancellor Friedrich Merz in South Korea, Japan and Singapore, Der Spiegel reported.
How it got here
Apple's warnings went to people in 99 countries.
Why it matters
European lawmaker Hannah Neumann says it looks like a large attack on Germany's political leaders.
How it works
Phone forensics experts separate being targeted from being infected, and the BSI says it cannot yet tell which is which.
Seen it before
A note saying someone tried your door handle does not tell you whether they got inside.
Nobody knows yet
How many German officials' phones were actually infected.
The BSI says it cannot yet tell infected devices from ones that were only targeted.
1 source
German lawmakers among Apple spyware targets
The full story
German lawmakers among Apple spyware targets
German lawmakers and foreign-policy staff were among people Apple warned last week about spyware, Germany's BSI says.
Why it matters. European lawmaker Hannah Neumann says it looks like a large attack on Germany's political leaders.
Harvard breach counted at 851,000 email addresses
Swipe up for the story ↑Scroll down for the story ↓
Data stolen from Harvard University in 2025 holds 851,000 unique email addresses, Have I Been Pwned says.
Have I Been Pwned is a service that tells people when their email turns up in a leak.
Alumni and donors
in the fundraising files
Parents and some students
also in the files
The files cover alumni, donors, parents, staff and some students, from its fundraising systems.
They include dates of birth, home addresses, ethnicity and citizenship status, The Register reports.
How it got here
18 Nov 2025
Harvard finds the break-in
Now
851,000 emails counted
Harvard found the break-in on 18 November 2025, and the ShinyHunters gang claimed it.
Why it matters
Harvard says the systems held no Social Security numbers, passwords or card details.
The truth
Harvard did not say how many people were hit, so this count comes from outside analysis of the leaked files.
Who pays
Harvard's donors, alumni and parents
Their dates of birth, addresses and citizenship status sit in files the gang published.
2 sources
Harvard breach counted at 851,000 email addresses
The full story
Harvard breach counted at 851,000 email addresses
Data stolen from Harvard University in 2025 holds 851,000 unique email addresses, Have I Been Pwned says.
Why it matters. Harvard says the systems held no Social Security numbers, passwords or card details.
School software breach reaches 1.3 million people
Swipe up for the story ↑Scroll down for the story ↓
Frontline Education, which makes office software for US schools, is telling over 1.3 million people their data was stolen.
14 Aug
hackers get in
Now
1.3 million people told
Hackers got in on 14 August through a flaw in another company's software, Frontline says.
Stolen data can include Social Security numbers, passport numbers and Medicaid IDs of staff and students.
State filings count 1,298,917 people, and the real total is likely higher, SecurityWeek reports.
Why it matters
Frontline says it has told school districts, which are helping it reach families by email or letter.
How it works
Security staff call this third-party risk: a hole in a supplier's product let attackers reach Frontline.
Seen it before
A burglar who gets in through a neighbour's shared fence never needs your front door.
Nobody knows yet
How many people the Frontline breach really reached.
The 1,298,917 figure comes from only five states' filings, and other states keep no public portal.
1 source
School software breach reaches 1.3 million people
The full story
School software breach reaches 1.3 million people
Frontline Education, which makes office software for US schools, is telling over 1.3 million people their data was stolen.
Why it matters. Frontline says it has told school districts, which are helping it reach families by email or letter.
Hackers push a backdoor through a security tool's updates
Swipe up for the story ↑Scroll down for the story ↓
Attackers used stolen logins to send a backdoor through the update system of cPGuard, a security suite for web servers.
02:06 UTC
poisoned updates begin
12 hours on
they are stopped
Servers that updated after 02:06 UTC on 2 November could get the poisoned versions.
The script took password hashes, database settings and server login keys.
Its maker, OpsShield, says the attacker took 32 files from each customer server.
Namecheap
a web hosting firm that offers it
Hosting firms use cPGuard, and Namecheap offers it on its servers.
Why it matters
OpsShield tells admins to update to 5.88.25 and change passwords, database logins and keys.
How it works
Security engineers call this a supply-chain attack: the attackers sent the backdoor through cPGuard's own update channel.
Seen it before
A fake parcel brought by the real postman gets opened, because you trust the postman.
Good news
OpsShield stopped the poisoned updates within 12 hours of the first one.
1 source
Hackers push a backdoor through a security tool's updates
The full story
Hackers push a backdoor through a security tool's updates
Attackers used stolen logins to send a backdoor through the update system of cPGuard, a security suite for web servers.
Why it matters. OpsShield tells admins to update to 5.88.25 and change passwords, database logins and keys.
Barracuda patches two flaws that give full control
Swipe up for the story ↑Scroll down for the story ↓
Barracuda, an email security company, told customers to patch two flaws that can give attackers full control of devices.
One weakness lets attackers change an email gateway's settings.
Both need access to the device's admin page, so Barracuda advises allowing only trusted networks there.
How it got here
In 2023 a hacking group called UNC4841 used another Barracuda gateway flaw for at least 7 months.
Why it matters
Almost a third of the gateways hacked in 2023 belonged to governments, Mandiant found.
How it works
Security engineers call this chaining: two weaker flaws used together give full control.
Seen it before
One key opens the gate and another opens the door, and with both you are inside.
1 source
Barracuda patches two flaws that give full control
The full story
Barracuda patches two flaws that give full control
Barracuda, an email security company, told customers to patch two flaws that can give attackers full control of devices.
Why it matters. Almost a third of the gateways hacked in 2023 belonged to governments, Mandiant found.
Ransomware hits UK schools in a suspected AI-run attack
Swipe up for the story ↑Scroll down for the story ↓
37 of 50
computers locked, of nearly 50 reached
An attacker using 3AM ransomware locked about 37 of nearly 50 computers it reached at a group of UK schools.
Bitdefender, a security firm, says the attacker may have used AI-written commands in an automatic loop.
The commands repeated with small changes every 40 to 100 seconds, which is what Bitdefender noticed.
The attacker took stored passwords from 3 backup servers and made 2 hidden admin accounts.
How it got here
10 Sep
the attack begins
19 Sep
the attack ends
The attack ran from 10 to 19 September.
Why it matters
The attacker also searched for links to other schools and joined separate networks together.
How it works
Security staff call this moving sideways: using passwords from one machine, here backup servers, to reach the next.
Seen it before
A thief who finds the caretaker's key ring can open every classroom.
The truth
Bitdefender says it cannot tell which AI model was used, or whether a person copied commands by hand.
Nobody knows yet
Which AI model, if any, ran the attack on the UK schools.
Bitdefender could not reach the attacker's machine, so it could only describe the pattern of commands.
1 source
Ransomware hits UK schools in a suspected AI-run attack
The full story
Ransomware hits UK schools in a suspected AI-run attack
An attacker using 3AM ransomware locked about 37 of nearly 50 computers it reached at a group of UK schools.
Why it matters. The attacker also searched for links to other schools and joined separate networks together.
Russia detains 3 in a rare hacking arrest
Swipe up for the story ↑Scroll down for the story ↓
Russian police detained 3 suspected members of VasyGrek, a gang that sent fake emails to steal company logins.
Suspected leader
house arrest
2 others
travel limits
The suspected leader is under house arrest, and 2 others face travel limits.
How it got here
The group, also called Fluffy Wolf, has targeted Russian companies since at least 2016.
Kaspersky
Russian security firm; helped
Kaspersky, a Russian security firm, helped the investigation, Russia's Interior Ministry said.
Why it matters
Such arrests are unusual in Russia, where many known hacking groups have worked for years.
How it works
Security staff call this phishing: here, an email dressed as a business partner's, with malware hidden in a finance file.
Seen it before
A letter with your bank's logo can still come from a stranger.
The truth
The ministry did not say whether police in other countries helped.
Good news
Russian police detained three suspected members of VasyGrek, a phishing gang that had targeted Russian companies since 2016.
1 source
Russia detains 3 in a rare hacking arrest
The full story
Russia detains 3 in a rare hacking arrest
Russian police detained 3 suspected members of VasyGrek, a gang that sent fake emails to steal company logins.
Why it matters. Such arrests are unusual in Russia, where many known hacking groups have worked for years.
Defence firm pays $13 million over cyber rules
Swipe up for the story ↑Scroll down for the story ↓
ILC Dover and ILC Astrospace, US defence and space suppliers, agreed to pay $13 million to settle US government claims.
Army, NASA, Missile Defense Agency
were billed
ILC
broke the rules, the US says
The US government says they billed the Army, NASA and the Missile Defense Agency while breaking contract security rules.
It says ILC kept sensitive data with a cloud firm without proper safeguards, from 2018 to 2024.
Ingersoll Rand
owner of ILC Dover since 2024
Parent company Ingersoll Rand says it inherited the problem when it bought ILC Dover in June 2024.
Why it matters
US Attorney Benjamin Wallace said the rules exist to stop bad actors taking advantage of US technical work.
How it works
Lawyers call this a False Claims Act case: billing the government while knowingly breaking the contract's terms.
Seen it before
Charging for a guarded car park while leaving the gate open is still charging for a guard.
The truth
ILC settled without admitting liability, and its owner says the problems were fixed by October 2024.
Who gains
Ingersoll Rand, owner of ILC Dover
Paying $13 million closes the US government's case without any admission of liability.
1 source
Defence firm pays $13 million over cyber rules
The full story
Defence firm pays $13 million over cyber rules
ILC Dover and ILC Astrospace, US defence and space suppliers, agreed to pay $13 million to settle US government claims.
Why it matters. US Attorney Benjamin Wallace said the rules exist to stop bad actors taking advantage of US technical work.
CrowdStrike buys two firms that check AI agents
Swipe up for the story ↑Scroll down for the story ↓
CrowdStrike, a big security company, agreed to buy two AI security firms, Fiddler AI and Dreadnode.
The Wall Street Journal put the prices at about $400 million and $95 million.
Fiddler watches AI agents, programs that act on their own, and explains the decisions they make.
Dreadnode attacks AI models with more than 100 known tricks before they go live.
“the flight recorder for every agent”
CrowdStrike's chief, George Kurtz, called Fiddler "the flight recorder for every agent".
What comes next
The deals are expected to close by the end of January 2027.
Why it matters
Agents can hold real logins and still make bad choices, which older security tests do not reach.
How it works
Engineers call this observability: keeping a record of each step a system takes, so a mistake can be traced.
Seen it before
A teacher who sees your working, not just the answer, can tell where you went wrong.
Who gains
CrowdStrike
Buying Fiddler and Dreadnode lets it sell checks on AI agents to the companies now putting agents to work.
2 sources
CrowdStrike buys two firms that check AI agents
The full story
CrowdStrike buys two firms that check AI agents
CrowdStrike, a big security company, agreed to buy two AI security firms, Fiddler AI and Dreadnode.
Why it matters. Agents can hold real logins and still make bad choices, which older security tests do not reach.
Android update fixes over 200 flaws
Swipe up for the story ↑Scroll down for the story ↓
Google's November Android updates fix more than 200 flaws, two dozen of them rated critical.
The worst lets an attacker run code on a phone without its owner tapping anything.
The first part fixes 119 flaws and the second part fixes 97.
Google sent 17 of the fixes early, through Google Play system updates.
Why it matters
Qualcomm
chip maker
MediaTek
chip maker
The second part covers parts made by Qualcomm, MediaTek and other chip makers inside many phones.
How it works
Google calls it the security patch level: a date that shows which fixes a phone holds.
Seen it before
A date stamp on milk tells you how fresh it is without opening it.
1 source
Android update fixes over 200 flaws
The full story
Android update fixes over 200 flaws
Google's November Android updates fix more than 200 flaws, two dozen of them rated critical.
Why it matters. The second part covers parts made by Qualcomm, MediaTek and other chip makers inside many phones.
NATO drill puts AI agents into cyber training
Swipe up for the story ↑Scroll down for the story ↓
More than 250 people from over 40 countries began NATO's Crossed Swords cyber exercise in Tallinn, Estonia, on Monday.
NATO's cyber defence centre
runs the drill
It is run by NATO's cyber defence centre in Tallinn and lasts until Friday.
New this year, players build AI agents and work beside them.
For the first time, the drill also uses OpenAI's Daybreak cyber defence tool.
Why it matters
Its director, Jozef Kostelansky, says cyber attacks must be timed to the minute with military action.
1 source
NATO drill puts AI agents into cyber training
The full story
NATO drill puts AI agents into cyber training
More than 250 people from over 40 countries began NATO's Crossed Swords cyber exercise in Tallinn, Estonia, on Monday.
Why it matters. Its director, Jozef Kostelansky, says cyber attacks must be timed to the minute with military action.
Russia-aligned spies target Ukrainian courts and councils
Swipe up for the story ↑Scroll down for the story ↓
ESET
security firm
Gorbag
Russia-aligned spy group
ESET, a security firm, says a Russia-aligned spy group it calls Gorbag has targeted Ukraine since February 2025.
Justice and police
about a quarter
Local councils
about a quarter
About a quarter of targets were justice and police bodies, and another quarter local councils near the front line.
Feb 2025
targeting begins
Early 2026
booby-trapped WinRAR files
In early 2026 it switched to booby-trapped files using a flaw in WinRAR, a common tool for opening compressed files.
Its stealer took saved passwords, browser cookies, VPN settings and documents.
Why it matters
ESET says the targets match Russia's wartime intelligence priorities, possibly including war-crimes cases.
How it works
Researchers call it a path traversal flaw: a file in the archive lands where it should not go.
Seen it before
A parcel addressed to the hall that ends up in your bedroom has got past the front door.
1 source
Russia-aligned spies target Ukrainian courts and councils
The full story
Russia-aligned spies target Ukrainian courts and councils
ESET, a security firm, says a Russia-aligned spy group it calls Gorbag has targeted Ukraine since February 2025.
Why it matters. ESET says the targets match Russia's wartime intelligence priorities, possibly including war-crimes cases.
Fake bank sites pretend to be shut down
Swipe up for the story ↑Scroll down for the story ↓
KELA, a security firm, found 248 live fake banking and scam sites tied to 189 US banks and financial firms.
On a second visit, the sites show a fake "account suspended" page.
Many of the site names pair a bank's name with words like secure or cancel.
Scanners and takedown teams check a reported site twice, so they see what looks like a finished takedown.
A victim usually follows a link only once, and sees the cloned bank page.
Why it matters
Many of the fake sites aim at business banking, KELA says.
How it works
Security staff call this cloaking: showing one page to checkers and another to victims.
Seen it before
A pupil who looks busy only when the teacher walks past is playing the same trick.
1 source
Fake bank sites pretend to be shut down
The full story
Fake bank sites pretend to be shut down
KELA, a security firm, found 248 live fake banking and scam sites tied to 189 US banks and financial firms.
Why it matters. Many of the fake sites aim at business banking, KELA says.
US agency confirms attacks on a WordPress flaw
Swipe up for the story ↑Scroll down for the story ↓
CISA, the US cyber-defence agency, added a WordPress core flaw to its list of bugs attackers are using.
WordPress
free software that runs many websites
WordPress is free software that runs many websites.
US federal agencies must fix flaws on that list quickly, under CISA's orders.
How it got here
A WordPress core race flaw fixed on 28 October drew nearly 63,000 attack attempts in 24 hours.
Why it matters
Website owners who have not updated stay open until they do.
How it works
Programmers call it a race condition: the attacker must act in the instant WordPress unpacks an update.
Seen it before
Grabbing a seat in the second it is empty is a race you win only by timing.
2 sources
US agency confirms attacks on a WordPress flaw
The full story
US agency confirms attacks on a WordPress flaw
CISA, the US cyber-defence agency, added a WordPress core flaw to its list of bugs attackers are using.
Why it matters. Website owners who have not updated stay open until they do.
Security investor Ballistic seeks $490 million
Swipe up for the story ↑Scroll down for the story ↓
Ballistic Ventures, a US investor that backs cybersecurity start-ups, is raising $490 million across two new funds, TechCrunch found.
Filings show $360 million for its third main fund and $130 million for a separate one.
Ted Schlein
co-founder; formerly of Kleiner Perkins
Ted Schlein, a former partner at the investor Kleiner Perkins, co-founded the San Francisco firm in late 2021.
It has backed Talon Cyber Security, later bought by Palo Alto Networks, and Pangea, bought by CrowdStrike.
How it got here
2022
first fund, $300 million
Now
$490 million sought
Its first fund, in 2022, was $300 million.
Neither new fund had taken in any money when the filings were made.
Why it matters
TechCrunch ties the raise to fears of AI-driven attacks and new openings as companies adopt AI agents.
1 source
Security investor Ballistic seeks $490 million
The full story
Security investor Ballistic seeks $490 million
Ballistic Ventures, a US investor that backs cybersecurity start-ups, is raising $490 million across two new funds, TechCrunch found.
Why it matters. TechCrunch ties the raise to fears of AI-driven attacks and new openings as companies adopt AI agents.
That's the whole Cyber day.
24 stories, 43 sources.