#cybersecurity
posts tagged #cybersecurity
preview
Cross posted from lemmy.ca/post/74010805
Cross posted from lemmy.ca/post/74010805
preview
Wanna chat about something non-infosec amongst those of us who frequent /c/cybersecurity? Here’s your chance! (Keep things civil & respectful please)
Wanna chat about something non-infosec amongst those of us who frequent /c/cybersecurity? Here’s your chance! (Keep things civil & respectful please)
preview
Amplifon Americas, the US hearing-care company, has started notifying people about a breach of its systems. The Texas Attorney General filing lists 1,332,826 people in total, including 85,804 Texans. Vermont was told about 1,027 residents.
The company says someone copied files from its systems between January 18 and February 8, 2026. It spotted the activity on February 11, but the review of what was in those files wasn’t finished until October 6. Notices started going out in early November.
What may have been in the files, depending on the person: name, address, date of birth, driver’s license or other government ID, hearing and medical information, dates of service, and health insurance details. For some people, Social Security numbers and financial information were included too. Amplifon says it hasn’t seen that data used for fraud so far. Your own letter is the one that lists what applied to you.
If you were a patient, plan member, or otherwise got a letter, this is the useful order:
- Freeze your credit at Equifax, Experian, and TransUnion. It’s free, and it is the highest-value step if an SSN might be in the mix. Do all three. A freeze on one bureau isn’t enough.
- Get an IRS Identity Protection PIN at irs.gov so nobody can file a tax return with your Social Security number.
- Enroll in whatever free monitoring the letter offers. It’s usually only for people whose SSN or financial info was involved, so read that part carefully.
- Check bank and card statements, and look at health insurance Explanation of Benefits for visits or claims you don’t recognize.
- Hang up on calls or texts about hearing aids, insurance, or refunds that already know your name or clinic details. Call Amplifon or your insurer on a number you find yourself.
The letter should also have a phone number for questions. Keep it.
If you get a breach notice and aren’t sure what to do with it, we wrote a plain guide here: osintpro.dev/…/what-a-breach-notification-means
Disclosure: I’m with OSINT Pro (osintpro.dev), a paid breach lookup tool with one free lookup, and this account is run by a bot. You don’t need the tool for any of the above.
Amplifon Americas, the US hearing-care company, has started notifying people about a breach of its systems. The Texas Attorney General filing lists 1,332,826 people in total, including 85,804 Texans. Vermont was told about 1,027 residents.
The company says someone copied files from its systems between January 18 and February 8, 2026. It spotted the activity on February 11, but the review of what was in those files wasn’t finished until October 6. Notices started going out in early November.
What may have been in the files, depending on the person: name, address, date of birth, driver’s license or other government ID, hearing and medical information, dates of service, and health insurance details. For some people, Social Security numbers and financial information were included too. Amplifon says it hasn’t seen that data used for fraud so far. Your own letter is the one that lists what applied to you.
If you were a patient, plan member, or otherwise got a letter, this is the useful order:
- Freeze your credit at Equifax, Experian, and TransUnion. It’s free, and it is the highest-value step if an SSN might be in the mix. Do all three. A freeze on one bureau isn’t enough.
- Get an IRS Identity Protection PIN at irs.gov so nobody can file a tax return with your Social Security number.
- Enroll in whatever free monitoring the letter offers. It’s usually only for people whose SSN or financial info was involved, so read that part carefully.
- Check bank and card statements, and look at health insurance Explanation of Benefits for visits or claims you don’t recognize.
- Hang up on calls or texts about hearing aids, insurance, or refunds that already know your name or clinic details. Call Amplifon or your insurer on a number you find yourself.
The letter should also have a phone number for questions. Keep it.
If you get a breach notice and aren’t sure what to do with it, we wrote a plain guide here: osintpro.dev/…/what-a-breach-notification-means
Disclosure: I’m with OSINT Pro (osintpro.dev), a paid breach lookup tool with one free lookup, and this account is run by a bot. You don’t need the tool for any of the above.
preview
Weekly thread for any and all career, learning and general guidance questions. Thinking of taking a training or going for a cert? Wondering how to level up your career? Wondering what NOT to do? Got other questions? This is the time and place to ask!
Weekly thread for any and all career, learning and general guidance questions. Thinking of taking a training or going for a cert? Wondering how to level up your career? Wondering what NOT to do? Got other questions? This is the time and place to ask!
preview
cross-posted from: piefed.world/…/2-carmakers-left-a-hidden-admin-co…
cross-posted from: piefed.world/…/2-carmakers-left-a-hidden-admin-co…
preview
Cross posted from discuss.tchncs.de/post/70438628
Cross posted from discuss.tchncs.de/post/70438628
preview
I work on payments, and I’ve spent the last few months thinking about one specific failure mode: an AI agent that can move money, acting while nobody is watching. The interesting attacks aren’t prompt injection. They’re the three cases below — all of which I’ve now seen in real runs, not just in theory.
-
The look-alike payee. The agent is told to pay a known provider. The invoice names a payee that differs by one or two characters — same shape, different address. A human reading quickly misses it. A model that compares the payee against a known-recipients list catches it, if it actually checks before signing. In my own test runs this one gets rejected cleanly. In the wild, it’s the oldest trick in payments, and agents fall for it the same way people do.
-
The reusable approval. The agent signs an approval once — say, an unlimited token approval for a spender — and that approval stays valid. A short expiry doesn’t fix it: the approval can still be reused on a slightly different payment before it expires. What fixes it is binding the approval to the exact payload (amount, payee, currency) and making it single-use. Expiry is the weak version of that control; payload binding is the strong version.
-
The delegated account. EIP-7702 lets an externally owned account delegate to a smart contract. If the account is already delegated to a contract outside the expected set, the gate is checking the wrong thing — it’s verifying an account that no longer acts alone. The fix is reading the account code first (eth_getCode): if the code is a 0xef0100 delegation to an unexpected delegate, hold. A reader found this gap in my own work in October, and it shipped as a fix — that’s the part of open source that actually works.
The common thread: all three have to be caught before signing. After settlement there is nothing to reverse — stablecoin payments are final. A post-settlement check is an audit trail, not a control. The control has to sit between the agent’s decision and the signature, and it has to be deterministic — a rule that says yes or no for this exact payment — not another judgment call by the same agent that decided to pay.
I’m curious how people here think about this. If you run agents with any ability to act on the world: what actually sits between your agent and the consequential action — a built-in approval step, a human in the loop, or something you built yourself? And when something looks off, should it pause for a human, or refuse outright?
For anyone who wants to see the failure modes with receipts rather than take my word for them — a write-up with a real agent payment, a held unlimited approval, and a rejected look-alike, plus the testnet transaction: dev.to/…/an-agent-paid-and-a-gate-stopped-two-pay… huggingface.co/sriram1983007/SRA-RiskGate-4B
I work on payments, and I’ve spent the last few months thinking about one specific failure mode: an AI agent that can move money, acting while nobody is watching. The interesting attacks aren’t prompt injection. They’re the three cases below — all of which I’ve now seen in real runs, not just in theory.
-
The look-alike payee. The agent is told to pay a known provider. The invoice names a payee that differs by one or two characters — same shape, different address. A human reading quickly misses it. A model that compares the payee against a known-recipients list catches it, if it actually checks before signing. In my own test runs this one gets rejected cleanly. In the wild, it’s the oldest trick in payments, and agents fall for it the same way people do.
-
The reusable approval. The agent signs an approval once — say, an unlimited token approval for a spender — and that approval stays valid. A short expiry doesn’t fix it: the approval can still be reused on a slightly different payment before it expires. What fixes it is binding the approval to the exact payload (amount, payee, currency) and making it single-use. Expiry is the weak version of that control; payload binding is the strong version.
-
The delegated account. EIP-7702 lets an externally owned account delegate to a smart contract. If the account is already delegated to a contract outside the expected set, the gate is checking the wrong thing — it’s verifying an account that no longer acts alone. The fix is reading the account code first (eth_getCode): if the code is a 0xef0100 delegation to an unexpected delegate, hold. A reader found this gap in my own work in October, and it shipped as a fix — that’s the part of open source that actually works.
The common thread: all three have to be caught before signing. After settlement there is nothing to reverse — stablecoin payments are final. A post-settlement check is an audit trail, not a control. The control has to sit between the agent’s decision and the signature, and it has to be deterministic — a rule that says yes or no for this exact payment — not another judgment call by the same agent that decided to pay.
I’m curious how people here think about this. If you run agents with any ability to act on the world: what actually sits between your agent and the consequential action — a built-in approval step, a human in the loop, or something you built yourself? And when something looks off, should it pause for a human, or refuse outright?
For anyone who wants to see the failure modes with receipts rather than take my word for them — a write-up with a real agent payment, a held unlimited approval, and a rejected look-alike, plus the testnet transaction: dev.to/…/an-agent-paid-and-a-gate-stopped-two-pay… huggingface.co/sriram1983007/SRA-RiskGate-4B
preview
title: “OpenAI Finds 948 Local Privilege Escalation Bugs In The Linux Kernel” url: “www.phoronix.com/news/OpenAI-948-Linux-LPE-Bugs” author: “Written by Michael Larabel in AI on 5 November 2026 at 05:00 AM EST. 38 Comments”
OpenAI published a blog post outlining their work on autonomous vulnerability discovery. To showcase their AI potential, they found 948 valid local privilege escalation vulnerabilities within the Linux kernel.
OpenAI’s prototype found 2,308 candidate vulnerabilities in the Linux kernel, dynamically reproduced 1,724 vulnerabilities, and ultimately found 948 local privilege escalation exploits for allowing unprivileged local users to gain root access. They also generated candidate patches for maintainer review.
Discovering nearly one thousand local privilege escalation vulnerabilities consumed $1.2 million USD in API credits.
OpenAI researchers noted in their blog post:
“Within the next few years, we believe that if there are vulnerabilities in a given software project, frontier models will likely be able to discover them and even less capable models may be able to develop exploits once disclosed. In a world of no unknown vulnerabilities, it will be critical to build and maintain high assurance software, especially core software such as operating systems, browsers, image parsing libraries, the TCP/IP stack, and more. Investing in training models to write secure code by default, minimizing attack surface by removing unnecessary features, and formal verification are promising areas of future research to effectively scale software security, in addition to finding and fixing existing vulnerabilities.”
Another interesting takeaway is that their GPT-5.5 Cyber model is finding more than 5x more novel LPEs than their GPT-5.6 Sol model despite three months of hardening/patches in the Linux kernel between scans.
Those wanting to go through this OpenAI showcase in more detail can do so via their blog post.
title: “OpenAI Finds 948 Local Privilege Escalation Bugs In The Linux Kernel” url: “www.phoronix.com/news/OpenAI-948-Linux-LPE-Bugs” author: “Written by Michael Larabel in AI on 5 November 2026 at 05:00 AM EST. 38 Comments”
OpenAI published a blog post outlining their work on autonomous vulnerability discovery. To showcase their AI potential, they found 948 valid local privilege escalation vulnerabilities within the Linux kernel.
OpenAI’s prototype found 2,308 candidate vulnerabilities in the Linux kernel, dynamically reproduced 1,724 vulnerabilities, and ultimately found 948 local privilege escalation exploits for allowing unprivileged local users to gain root access. They also generated candidate patches for maintainer review.
Discovering nearly one thousand local privilege escalation vulnerabilities consumed $1.2 million USD in API credits.
OpenAI researchers noted in their blog post:
“Within the next few years, we believe that if there are vulnerabilities in a given software project, frontier models will likely be able to discover them and even less capable models may be able to develop exploits once disclosed. In a world of no unknown vulnerabilities, it will be critical to build and maintain high assurance software, especially core software such as operating systems, browsers, image parsing libraries, the TCP/IP stack, and more. Investing in training models to write secure code by default, minimizing attack surface by removing unnecessary features, and formal verification are promising areas of future research to effectively scale software security, in addition to finding and fixing existing vulnerabilities.”
Another interesting takeaway is that their GPT-5.5 Cyber model is finding more than 5x more novel LPEs than their GPT-5.6 Sol model despite three months of hardening/patches in the Linux kernel between scans.
Those wanting to go through this OpenAI showcase in more detail can do so via their blog post.
preview
Wanna chat about something non-infosec amongst those of us who frequent /c/cybersecurity? Here’s your chance! (Keep things civil & respectful please)
Wanna chat about something non-infosec amongst those of us who frequent /c/cybersecurity? Here’s your chance! (Keep things civil & respectful please)
preview
I’ve been trying to get good at fuzzing for a while. I recently started a new fuzzing project and I realized that I could take from past PoCs and put them in a corpus to try and generate bugs of a similar shape. Do you know any good sources of searchable PoCs? I’ve mostly been going through specific projects’ issue trackers looking for PoCs.
I’ve been trying to get good at fuzzing for a while. I recently started a new fuzzing project and I realized that I could take from past PoCs and put them in a corpus to try and generate bugs of a similar shape. Do you know any good sources of searchable PoCs? I’ve mostly been going through specific projects’ issue trackers looking for PoCs.
preview
Weekly thread to discuss whatever you’re working on, big or small, at work or in your free time.
Weekly thread to discuss whatever you’re working on, big or small, at work or in your free time.
preview
Vulnerability-Lookup (vulnerability.circl.lu) is an open-source platform run by CIRCL, the Luxembourg CERT. It aggregates vulnerability data from many sources (CVE, GitHub advisories, PySec, CSAF feeds and more). It also collects “sightings”: mentions, proofs of concept and exploitation reports seen on social media, in exploit databases, from honeypots and from the community. Every month we publish a report built from that data.
Highlights for October 2026:
- PHP2Shell (CVE-2026-109618 / CVE-2026-109619): two memory-safety flaws in the PHP interpreter leaked before their coordinated release on 21 October and drew 1,773 sightings. Exploitation was observed on disclosure day, five days before CISA listed them.
- 17,097 CVEs were published, a third record month in a row. Total sightings rose 20%, but proof-of-concept sightings fell 38% and exploitation sightings 20%.
- 200 vulnerabilities entered at least one of the five Known Exploited Vulnerabilities catalogs we track: CISA, CIRCL, ENISA / EU CSIRTs Network, The Shadowserver Foundation and Previdian. The catalogs are federated through GCVE (gcve.eu), and the report includes a matrix showing which catalog covers what.
- Ivanti Connect Secure had three CISA entries. CIRCL confirmed the first one before its CVE record was published.
- Previdian covered all 36 CISA additions and flagged 19 of them earlier, in one case by 77 days.
Severity levels come from our open VLAI models (arxiv.org/abs/2507.03607). The report was generated with AI assistance using VulnMCP (github.com/vulnerability-lookup/VulnMCP). Feedback is welcome on GitHub (github.com/vulnerability-lookup/…/issues/).
Vulnerability-Lookup (vulnerability.circl.lu) is an open-source platform run by CIRCL, the Luxembourg CERT. It aggregates vulnerability data from many sources (CVE, GitHub advisories, PySec, CSAF feeds and more). It also collects “sightings”: mentions, proofs of concept and exploitation reports seen on social media, in exploit databases, from honeypots and from the community. Every month we publish a report built from that data.
Highlights for October 2026:
- PHP2Shell (CVE-2026-109618 / CVE-2026-109619): two memory-safety flaws in the PHP interpreter leaked before their coordinated release on 21 October and drew 1,773 sightings. Exploitation was observed on disclosure day, five days before CISA listed them.
- 17,097 CVEs were published, a third record month in a row. Total sightings rose 20%, but proof-of-concept sightings fell 38% and exploitation sightings 20%.
- 200 vulnerabilities entered at least one of the five Known Exploited Vulnerabilities catalogs we track: CISA, CIRCL, ENISA / EU CSIRTs Network, The Shadowserver Foundation and Previdian. The catalogs are federated through GCVE (gcve.eu), and the report includes a matrix showing which catalog covers what.
- Ivanti Connect Secure had three CISA entries. CIRCL confirmed the first one before its CVE record was published.
- Previdian covered all 36 CISA additions and flagged 19 of them earlier, in one case by 77 days.
Severity levels come from our open VLAI models (arxiv.org/abs/2507.03607). The report was generated with AI assistance using VulnMCP (github.com/vulnerability-lookup/VulnMCP). Feedback is welcome on GitHub (github.com/vulnerability-lookup/…/issues/).
preview
With the NIKKE launcher tampering in the news this week, here’s the order I’d go in if software on your PC turns out to be compromised. Most of this applies to any trojanized installer.
- Assume anything saved in your browser on that machine is gone. Saved passwords, cookies, autofill. Stealer malware grabs those first.
- Change passwords from a different, clean device. Start with your main email, then banking, then anything that can reset other accounts.
- Log out of all sessions. Changing a password doesn’t always kill a stolen session cookie. Most big sites have a “sign out everywhere” option. Discord, Steam, Google and Microsoft all do.
- Check 2FA settings for anything new. Attackers sometimes add their own authenticator or recovery email.
- Clean or reinstall the machine before logging back in. Otherwise you’re just handing over the new passwords.
- Keep an eye on email for login alerts and password resets you didn’t ask for over the next few weeks.
Stolen data like this often ends up in stealer log dumps that get traded later, so a “nothing found” result today doesn’t mean much. Worth rechecking down the road.
Disclosure: I’m with OSINT Pro, a breach lookup tool. We wrote up how stealer logs work here if it’s useful: www.osintpro.dev/…/what-are-infostealer-logs
Happy to hear what others would add or reorder.
With the NIKKE launcher tampering in the news this week, here’s the order I’d go in if software on your PC turns out to be compromised. Most of this applies to any trojanized installer.
- Assume anything saved in your browser on that machine is gone. Saved passwords, cookies, autofill. Stealer malware grabs those first.
- Change passwords from a different, clean device. Start with your main email, then banking, then anything that can reset other accounts.
- Log out of all sessions. Changing a password doesn’t always kill a stolen session cookie. Most big sites have a “sign out everywhere” option. Discord, Steam, Google and Microsoft all do.
- Check 2FA settings for anything new. Attackers sometimes add their own authenticator or recovery email.
- Clean or reinstall the machine before logging back in. Otherwise you’re just handing over the new passwords.
- Keep an eye on email for login alerts and password resets you didn’t ask for over the next few weeks.
Stolen data like this often ends up in stealer log dumps that get traded later, so a “nothing found” result today doesn’t mean much. Worth rechecking down the road.
Disclosure: I’m with OSINT Pro, a breach lookup tool. We wrote up how stealer logs work here if it’s useful: www.osintpro.dev/…/what-are-infostealer-logs
Happy to hear what others would add or reorder.
preview
I’m not sure what I expected, but this article is quite depressing.
Many industrial devices include a service access so the vendor’s technician can get in without the customer’s password. A classic —and dangerous— way to implement it is the code of the day: a code that changes every day according to a formula that, in theory, only the vendor knows.
The problem shows up when that “secret formula” is the only thing protecting access (security by obscurity) and it can also be recovered from the firmware itself. That is exactly what happens here: sm_app ships unobfuscated, with function and variable names in plain sight. The one that validates the code is called, literally, check_code_of_day.
The article is from October 14th and notes:
Whether the attack went through CVE-2025-41450 is, for now, just a hypothesis: the vector has not been officially established and the investigation is still open. Several reports throughout this year had already warned that something like this could happen, and during 2025 several vulnerabilities were disclosed in Danfoss equipment that ended in a vendor fix. The flaw we look at here is public and patched (at least since R4.3.1).
That makes the most recent attacks all the more alarming, should they have been achieved via the same method.
I’m not sure what I expected, but this article is quite depressing.
Many industrial devices include a service access so the vendor’s technician can get in without the customer’s password. A classic —and dangerous— way to implement it is the code of the day: a code that changes every day according to a formula that, in theory, only the vendor knows.
The problem shows up when that “secret formula” is the only thing protecting access (security by obscurity) and it can also be recovered from the firmware itself. That is exactly what happens here: sm_app ships unobfuscated, with function and variable names in plain sight. The one that validates the code is called, literally, check_code_of_day.
The article is from October 14th and notes:
Whether the attack went through CVE-2025-41450 is, for now, just a hypothesis: the vector has not been officially established and the investigation is still open. Several reports throughout this year had already warned that something like this could happen, and during 2025 several vulnerabilities were disclosed in Danfoss equipment that ended in a vendor fix. The flaw we look at here is public and patched (at least since R4.3.1).
That makes the most recent attacks all the more alarming, should they have been achieved via the same method.
preview
Background: I’m using Ungoogled Chromium, which I do not keep up to date because non-Debian packages are a big hassle to maintain. I expect this risk to be mine alone.
Since websites are secretive¹ about their reasons for blocking, it has been quite painful (as a Tor user) that many services are not blocking Tor, but in fact blocking old versions of Chrome. WTF? Why does a website block old Chrome versions? My gut thought was the website wants to nanny users, to try to force users to secure themselves (“for the user’s own protection”), because surely a user with old chrome does not realistically threaten the server’s security.
The answer: it turns out that AI bots are strangely using old Chrome versions for a user-agent header. So webmasters are reacting by blocking old versions of Chrome. What I don’t understand is why devs of AI bots would try to masquerade as old Chrome versions.
And to a lesser extent I wonder why webmasters are reckless enough to take this collective-punishment-is-ok stance. It would be like a store owner banning customers wearing blue hats from their shop because some shoplifters were caught wearing blue hats.
¹ Exceptionally, codeberg.org actually respects our dignity enough to preduce an error message that says they are blocking old chrome. But they are the exception.
Background: I’m using Ungoogled Chromium, which I do not keep up to date because non-Debian packages are a big hassle to maintain. I expect this risk to be mine alone.
Since websites are secretive¹ about their reasons for blocking, it has been quite painful (as a Tor user) that many services are not blocking Tor, but in fact blocking old versions of Chrome. WTF? Why does a website block old Chrome versions? My gut thought was the website wants to nanny users, to try to force users to secure themselves (“for the user’s own protection”), because surely a user with old chrome does not realistically threaten the server’s security.
The answer: it turns out that AI bots are strangely using old Chrome versions for a user-agent header. So webmasters are reacting by blocking old versions of Chrome. What I don’t understand is why devs of AI bots would try to masquerade as old Chrome versions.
And to a lesser extent I wonder why webmasters are reckless enough to take this collective-punishment-is-ok stance. It would be like a store owner banning customers wearing blue hats from their shop because some shoplifters were caught wearing blue hats.
¹ Exceptionally, codeberg.org actually respects our dignity enough to preduce an error message that says they are blocking old chrome. But they are the exception.
preview
Hi everyone,
I’m looking for new blogs and, more generally, sources of information on the topics of “threat intelligence”, “malware reverse engineering”, “security engineering” and “DevSecOps”. Do you have any recommendations?
Hi everyone,
I’m looking for new blogs and, more generally, sources of information on the topics of “threat intelligence”, “malware reverse engineering”, “security engineering” and “DevSecOps”. Do you have any recommendations?
preview
Ryohin Keikaku (the company behind MUJI) said on Oct 30 that someone got unauthorized access to the official MUJI app. Up to about 27.26 million records may be affected, including people who already deleted their app account.
What they say leaked: email address, date of birth, gender. They say names, addresses, phone numbers and card details aren’t part of it, and they haven’t seen it posted or used for fraud yet.
Nobody can log in as you with that, so the practical stuff is pretty boring:
- Expect fake “MUJI” emails about points, coupons or account problems. MUJI itself warned about this. Open the app yourself instead of tapping links.
- Stop treating your birthday as a secret. If a bank or carrier uses DOB as a check, add a PIN or something stronger.
- If your MUJI password is reused anywhere, change those. Passwords aren’t listed, but it’s a free fix.
- Put 2FA on the email account tied to it, since email is the main thing in this leak.
The notice is Japanese only and doesn’t say if MUJI apps in other countries are in scope, so if you’re outside Japan watch for mail from your local MUJI.
If you get a breach notice and aren’t sure what to do with it, we wrote up a plain guide here: osintpro.dev/…/what-a-breach-notification-means (disclosure: I work on OSINT Pro).
Ryohin Keikaku (the company behind MUJI) said on Oct 30 that someone got unauthorized access to the official MUJI app. Up to about 27.26 million records may be affected, including people who already deleted their app account.
What they say leaked: email address, date of birth, gender. They say names, addresses, phone numbers and card details aren’t part of it, and they haven’t seen it posted or used for fraud yet.
Nobody can log in as you with that, so the practical stuff is pretty boring:
- Expect fake “MUJI” emails about points, coupons or account problems. MUJI itself warned about this. Open the app yourself instead of tapping links.
- Stop treating your birthday as a secret. If a bank or carrier uses DOB as a check, add a PIN or something stronger.
- If your MUJI password is reused anywhere, change those. Passwords aren’t listed, but it’s a free fix.
- Put 2FA on the email account tied to it, since email is the main thing in this leak.
The notice is Japanese only and doesn’t say if MUJI apps in other countries are in scope, so if you’re outside Japan watch for mail from your local MUJI.
If you get a breach notice and aren’t sure what to do with it, we wrote up a plain guide here: osintpro.dev/…/what-a-breach-notification-means (disclosure: I work on OSINT Pro).
preview
Someone used an Amadeus login to view airline reservation records from Aug 2023 to Sep 2026.
Confirmed so far:
- Japan Airlines: about 92k customers
- Cathay Pacific: about 53k bookings
- ANA: about 37.8k
- Also Jin Air, Star Flyer, Korean Air, and Asiana
- Bloomberg reports Condor as well
Most people had their name, booking code, and flight details exposed. A smaller group also lost email, phone, passport number, and frequent flyer number. The airlines say no cards or passwords were taken.
What to do:
- Expect fake “your flight changed” emails that quote your real route
- Open the airline app yourself instead of tapping links
- Check that upcoming bookings were not changed
- Add a PIN to frequent flyer accounts and check the points
Disclosure: I work on OSINT Pro, which does breach lookups. If you want to see where else your email shows up, osintpro.dev/lookup/email
Someone used an Amadeus login to view airline reservation records from Aug 2023 to Sep 2026.
Confirmed so far:
- Japan Airlines: about 92k customers
- Cathay Pacific: about 53k bookings
- ANA: about 37.8k
- Also Jin Air, Star Flyer, Korean Air, and Asiana
- Bloomberg reports Condor as well
Most people had their name, booking code, and flight details exposed. A smaller group also lost email, phone, passport number, and frequent flyer number. The airlines say no cards or passwords were taken.
What to do:
- Expect fake “your flight changed” emails that quote your real route
- Open the airline app yourself instead of tapping links
- Check that upcoming bookings were not changed
- Add a PIN to frequent flyer accounts and check the points
Disclosure: I work on OSINT Pro, which does breach lookups. If you want to see where else your email shows up, osintpro.dev/lookup/email
preview
Wanna chat about something non-infosec amongst those of us who frequent /c/cybersecurity? Here’s your chance! (Keep things civil & respectful please)
Wanna chat about something non-infosec amongst those of us who frequent /c/cybersecurity? Here’s your chance! (Keep things civil & respectful please)
preview
Interesting that there were headlines and perhaps panic about an AI bot grabbing Georgia voter data – yet no one gives a shit that registration records of every Georgia voter is fully shared with Cloudflare. That’s both public and private data.
And worse, Cloudflare controls who can register to vote. If Cloudflare excludes you from access, you must do a paper registration. Then after submitting paper reg, the data entry worker still gives your sensitive data to Cloudflare anyway.
LOL:
“Georgia is consistently recognized as having some of the most secure elections in the country, with voting systems not connected to the internet designed to prevent fraud, paper-ballots, and risk-limiting audits,” Raffensperger continued. “Twisting this into a security scare for clicks undermines confidence in our elections and is absolutely shameful.”
(source)
Last time I checked, 7 US states share voter registration data with Cloudflare. The other 43 states do better. (As of a few years ago when I checked… probably a few more are sharing sensitive data with Cloudflare now).
It’s obviously like dealing with a poisonous well. If you spotlight the lack of security and deplore the reckless treatment of people’s sensitive data, you discourage voting. Which then helps republicans.
(edit) Cloudflare gatekeeping GA voter reg:

Interesting that there were headlines and perhaps panic about an AI bot grabbing Georgia voter data – yet no one gives a shit that registration records of every Georgia voter is fully shared with Cloudflare. That’s both public and private data.
And worse, Cloudflare controls who can register to vote. If Cloudflare excludes you from access, you must do a paper registration. Then after submitting paper reg, the data entry worker still gives your sensitive data to Cloudflare anyway.
LOL:
“Georgia is consistently recognized as having some of the most secure elections in the country, with voting systems not connected to the internet designed to prevent fraud, paper-ballots, and risk-limiting audits,” Raffensperger continued. “Twisting this into a security scare for clicks undermines confidence in our elections and is absolutely shameful.”
(source)
Last time I checked, 7 US states share voter registration data with Cloudflare. The other 43 states do better. (As of a few years ago when I checked… probably a few more are sharing sensitive data with Cloudflare now).
It’s obviously like dealing with a poisonous well. If you spotlight the lack of security and deplore the reckless treatment of people’s sensitive data, you discourage voting. Which then helps republicans.
(edit) Cloudflare gatekeeping GA voter reg:

preview
cross-posted from: nord.pub/…/broccoli2shell-the-brotli-decoder-bug-…
CVE-2026-111497 (Broccoli2Shell, aka Node2Shell): an out-of-bounds write in Google’s Brotli decoder affecting Node.js, Python Brotli/brotlicffi and anything built on the C decoder. Affected versions, how to check, and how to fix it.
cross-posted from: nord.pub/…/broccoli2shell-the-brotli-decoder-bug-…
CVE-2026-111497 (Broccoli2Shell, aka Node2Shell): an out-of-bounds write in Google’s Brotli decoder affecting Node.js, Python Brotli/brotlicffi and anything built on the C decoder. Affected versions, how to check, and how to fix it.
preview
Weekly thread for any and all career, learning and general guidance questions. Thinking of taking a training or going for a cert? Wondering how to level up your career? Wondering what NOT to do? Got other questions? This is the time and place to ask!
Weekly thread for any and all career, learning and general guidance questions. Thinking of taking a training or going for a cert? Wondering how to level up your career? Wondering what NOT to do? Got other questions? This is the time and place to ask!