Category: Hacks and Exploits

  • Update Outlook Now: Attackers Are Already Using This Flaw

    Update Outlook Now: Attackers Are Already Using This Flaw

    Opening one bad file in Outlook can leak your login to an attacker, and Microsoft says that’s already happening. I’d install the fix before you open another file someone sends you.

    What’s wrong

    On Nov 10, Microsoft released a fix for a flaw in Outlook and marked it as already being used in real attacks. The same day, CISA (the federal government’s cybersecurity agency) added it to its list of flaws known to be used by attackers.

    Here’s how it works. Someone sends you a specially made message file and talks you into opening it in Outlook. When you open it, Outlook connects to a computer the attacker controls, and it skips the warning it normally shows before loading anything from outside. That connection can leak the scrambled login information Windows uses to prove who you are, and the attacker can use it to sign in as you.

    Seeing the email in Outlook’s preview pane does not trigger it. You have to open the file. The attacker doesn’t need to sign in to anything to try this.

    How to fix it

    A note if you use new Outlook: Microsoft’s list of affected versions covers classic Outlook in the Office desktop apps, not new Outlook for Windows. New Outlook is a separate app, and it checks for and installs its own updates when you open it. The steps below are for classic Outlook and the other Office apps.

    Install the November 2026 Office security update. These are Microsoft’s steps for updating Office on Windows:

    1. Open any Office app, such as Word, and create a new document.
    2. Go to File > Account. If you opened classic Outlook, it’s File > Office Account.
    3. Under Product Information, choose Update Options > Update Now. If you don’t see Update Now, click Enable Updates first.
    4. When you see “You’re up to date!” or “Updates were installed,” close that window.

    If you only see an About button and no Update Options, your company manages your Office updates. Ask your IT help desk to install the November 2026 Office security update. You can also try Microsoft Update, the Windows Update setting that also installs Office updates.

    Microsoft’s full guide is here: How to update Microsoft 365 or Office for Windows

    More info

    Microsoft calls this the “Microsoft Outlook Spoofing Vulnerability,” and its CVE number is CVE-2026-124194 (a public ID that security folks use to track a specific flaw). Spoofing means pretending to be someone else, which is what the attacker gets to do once they have your login information.

    Microsoft rates it Moderate, with a severity score of 6.5 out of 10. Don’t let “Moderate” fool you. Microsoft and CISA both say it is already being used in attacks, and Microsoft lists it as needing action from you, which means installing the update.

    The flaw had not been made public before Microsoft released the fix.

    The login information that leaks is called an NTLMv2 hash (a scrambled version of your Windows sign-in that Windows sends to prove who you are). Normally Outlook warns you before it reaches out to an outside location. This flaw lets a crafted file skip that warning.

    Microsoft lists these versions as affected: Microsoft 365 Apps for Enterprise, Office LTSC 2021 and Office LTSC 2024 (LTSC is the version businesses buy once instead of by subscription). Microsoft has not shared who is behind the attacks or how many people were hit.

    You can read Microsoft’s own write-up on its CVE-2026-124194 page and its Microsoft 365 Apps security update release notes, the public CVE.org record, and CISA’s entry in its Known Exploited Vulnerabilities catalog.

  • Update Every Apple Device You Own Today, ASAP

    Update Every Apple Device You Own Today, ASAP

    I highly recommend you update every Apple device you own, today. On November 9, 2026, Apple released nine security updates, including ones for iPhone, iPad, Mac and Apple Watch. I’d rather you get ahead of this one, so here is what’s wrong, here is how to fix it, and if you care to learn more, the deeper details are at the bottom.

    Here’s what’s wrong

    The two fixes that got my attention are both Bluetooth pairing bugs. Pairing is the step where two devices agree to connect, like your phone and your earbuds.

    • With the first bug, a malicious Bluetooth device may be able to pair with yours without you tapping anything.
    • With the second, someone within Bluetooth range may be able to get around the limit on how many times they can guess the pairing code, then pair with your device without you tapping anything.

    Apple also fixed a few related problems:

    • AirPlay (Apple’s tool for sending video and audio to a TV or speaker) had the same kind of code-guessing problem. One bug could be tried from Bluetooth range, and a separate one from the same network.
    • On most devices, two iMessage group-chat bugs are fixed. With one, a stranger may be able to join a group chat and get its messages. With the other, someone added to a group may be able to read messages sent before they joined.

    Here’s how to fix it

    Updating is the fix, so update everything.

    On an iPhone or iPad, go to Settings, then General, then Software Update. On a Mac, go to System Settings, then General, then Software Update. On an Apple Watch, open Settings on the watch, then General, then Software Update.

    Want to learn more?

    You can skip this part. Updating is what matters, and this section is just for anyone who wants the details. Every flaw below has a CVE number (a public ID that security folks use to track a specific flaw). The quotes are Apple’s own wording from its security notes.

    The two Bluetooth pairing bugs are in CoreUtils (a behind-the-scenes part of Apple’s software), and both are fixed in all nine updates.

    • CVE-2026-113965, found by Apple: “A malicious Bluetooth device may be able to establish a pairing without user interaction.” Apple says it fixed this “with improved checks.”
    • CVE-2026-113964, found by an anonymous researcher: “An attacker in Bluetooth range may be able to bypass the rate limit on pairing code attempts and pair with a device without user interaction.” A rate limit is the cap on how many guesses are allowed. Apple calls this a logic issue, fixed “with improved state management.”

    The other fixes:

    • Bluetooth, CVE-2026-113970, in all nine updates: “An attacker in Bluetooth range may be able to corrupt process memory.” Apple calls it a use-after-free issue (a mistake where software keeps using a piece of memory after it has let it go).
    • AirPlay, CVE-2026-113968, in all nine updates: “An attacker in Bluetooth range may be able to bypass the rate limit on pairing code attempts.”
    • rapportd (a background Apple service), CVE-2026-113969, in all nine updates: “An attacker on the same network may be able to bypass the rate limit on AirPlay pairing code attempts.”
    • Messages, CVE-2026-113961: “A remote attacker may be able to join an iMessage group chat and receive its messages.”
    • Messages, CVE-2026-113960: “A user added to an iMessage group may be able to access messages sent before they joined.” Both Messages fixes are in every update except iOS 18.7.12 and watchOS 26.6.1.
    • App Intents (the part that lets apps hand actions to Siri and Shortcuts), CVE-2026-113963, in iOS and iPadOS 27.2.1, watchOS 27.2.1 and visionOS 27.2.1 only: “An app may be able to access sensitive user data.”
    • brotli (a tool that shrinks data so it moves faster), CVE-2026-111497, in every update except iOS 18.7.12 and watchOS 26.6.1: “Processing maliciously crafted input may lead to memory corruption.”
    • MediaRemote (a part of iOS and iPadOS), CVE-2026-113971, in iOS and iPadOS 18.7.12 only: “An attacker with physical access may be able to pair a device with an AirPlay receiver from the lock screen.”

    You can read Apple’s full list here: https://support.apple.com/en-us/100100

  • cPanel Patched a Root Flaw That Let Any Hosting Account Take the Whole Server

    cPanel Patched a Root Flaw That Let Any Hosting Account Take the Whole Server

    cPanel fixed three security flaws, and the ugly one is simple: any logged-in hosting account could run code as root and take full control of the server. Another flaw could modify other accounts’ databases.

    If you have a site on shared hosting, that is not a nerdy footnote. That is a neighbor on the same machine getting the master key to the building.

    What cPanel actually patched

    Three holes got closed. The headline one lets a regular cPanel login execute code with root privileges. Root is not a fancy admin setting. Root is the operating system treating you as the owner of the box. Sites, mail, files, configs, other customers. All of it.

    The other named problem is almost as bad for anyone who stores real data on that server. An account could change other accounts’ databases. On shared hosting, that is other people’s stores, member lists, and the content sitting next door to you on the same hardware.

    cPanel is still the dashboard a huge chunk of the everyday web runs on. Cheap shared plans, small shop sites, church pages, the agency that sold you a domain and a login. If you have ever clicked those little icons for email, files, and databases, you are in this neighborhood.

    Why root on a shared server is the whole ballgame

    Shared hosting only works if the walls between accounts hold. Those walls are software. When a logged-in account can become root, isolation is a sales pitch, not a fact.

    Full server control means one cheap hosting login could:

    • Read or change every site on that machine
    • Steal databases, passwords, and anything stored next to you
    • Plant junk that gets served from every domain on the box
    • Use the server as a launch pad for the next mess

    You did not have to be the target. You just had to rent space on the same server as someone who was.

    What this means if you only rent a hosting account

    Most regular people do not patch cPanel. Your host does. Or they do not.

    That is the part that pisses me off. You pay every month and assume the company with the keys is doing the adult work. Plenty of hosts are slow. Plenty of resellers are asleep. The flaw is in the panel. The exposure is whoever is still running the unpatched version.

    If you run your own VPS or dedicated box with cPanel, this one is on you. Update it. Do not wait for a convenient window three weeks from now.

    What you should do this week

    • If you use shared cPanel hosting, ticket your host and ask if the latest cPanel security updates are installed. Get a yes in writing.
    • If you run cPanel yourself, apply the vendor updates now and confirm services came back clean.
    • Look at your sites for odd admin users, mystery plugins, and database changes you did not make.
    • Keep backups you control, not only copies sitting on the same server that just became a root target.
    • If a host cannot tell you whether they patched, that is your answer. Start looking for a grown-up.

    This is a systems problem. Isolation either works or it does not. When a hosting account can become root, it does not. The patch is out. The only question left is whether the person who actually runs your server put it on.

    Primary Source: https://thehackernews.com/2026/09/new-cpanel-flaw-lets-hosting-account_0272795595.html

  • Hackers Used WordPress Flaws to Steal 18,566 Government Records and Passwords in Plain Text

    Hackers Used WordPress Flaws to Steal 18,566 Government Records and Passwords in Plain Text

    Chinese-speaking hackers exploited WordPress flaws to breach 49 organizations, steal 18,566 government records, grab credentials, and plant webshells. They also walked off with passwords stored in plaintext. That is the picture from a September 22, 2026 report, and it is ugly in a very ordinary way.

    Nobody needed a genius crew for this. They needed WordPress left open and passwords sitting where anyone who got in could read them like a grocery list.

    49 organizations, government records, and passwords in the clear

    Here is the haul in plain numbers. Attackers hit 49 organizations through WordPress flaws. They stole 18,566 government records. They stole credentials. They planted webshells so they could come back later. And they found passwords in plaintext, which means the logins were readable as typed, not scrambled.

    WordPress is not some obscure tool. It is the website software a huge share of regular outfits actually use because it is cheap, familiar, and easy to stand up. Easy to stand up is also easy to neglect. Plugins rot. Core goes unpatched. Some shop or office stores logins the same way you would write them on a sticky note. That is how government records end up in someone else’s hands.

    Plaintext passwords are a systems failure

    If a site is keeping passwords in plaintext, the rest of the security talk is theater. Once those logins are copied, they get tried on email, payroll, vendor portals, and personal accounts because people reuse the same password everywhere. Mix in 18,566 government records and this stops being an IT headache. It is other people’s information, sitting behind a WordPress site that never got treated like it mattered.

    Webshells mean they meant to stay

    A webshell is a spare key under the mat. The first hole gets them in. The shell lets them return, poke around, and pull more later. Patching the original WordPress flaw and calling it done is not a cleanup if the backdoor is still on the server.

    Anyone running these sites for a town office, a school, a clinic, or a small business needs to hear that part. The break-in is the start of the problem, not the end of it.

    What you should do if you run WordPress

    This is not a lecture for a corporate security team. This is for the person who actually keeps the website alive.

    • Update WordPress core, themes, and every plugin. Old code is the front door in this story.
    • Stop storing passwords in plaintext. If a plugin or a custom bit of the site does that, fix it or rip it out.
    • Lock down admin logins with unique passwords and two-factor authentication. Reused passwords are how a website breach becomes an email and bank problem.
    • After any suspected break-in, hunt for webshells and files that do not belong. Do not assume a patch closed the whole mess.
    • If you put government or customer records on WordPress, treat backups, access, and updates like a real job, not a weekend chore.

    Regular people did not volunteer to have their government records sitting on a neglected website. The 49 organizations that got hit had one job: keep the door shut. They did not. That should make you mad, and it should make you check your own site before someone else does.

    Primary Source: https://cybersecuritynews.com/wordpress-flaws/

  • ChatGPT Now Follows You Onto Other Websites With OpenAI’s One-Year Ad Cookie

    ChatGPT Now Follows You Onto Other Websites With OpenAI’s One-Year Ad Cookie

    OpenAI is connecting what you do on ordinary websites to your ChatGPT account. Researcher Buchodi reproduced the mechanism on a phone, verified it with two independent capture methods, and cross-checked months of traffic covering 936 distinct advertiser pixels across 1,029 hostnames.

    Here is the plain version. Companies that buy ads on ChatGPT drop a piece of OpenAI tracking code on their own sites, the same way they already drop Meta and Google pixels. That code sends a cookie called __obi back to OpenAI along with data about the page you are on. Products you search. Articles you read. Purchase behavior. Then OpenAI can join that activity to the ChatGPT identity sitting on your account.

    How the __obi cookie gets on your device

    It starts on chatgpt.com. The client generates 16 random bytes and posts them to OpenAI’s backend. The backend hands back a short-lived signed token, an RS256 JWT that expires in 60 seconds. That token binds your account identifier to a value called obi. OpenAI’s internal name for the ads platform is bazaar, which is why the collector lives at bzr.openai.com. The issuing service is called wadi.

    The client then posts that token cross-site to bzr.openai.com. The response sets the cookie:

    • Name: __obi
    • Domain: .openai.com
    • HttpOnly and Secure
    • SameSite=None, which is the setting that lets a cookie travel on cross-site requests
    • Max-Age: 31536000 seconds, which is one year

    That cookie value matches the obi value inside the token. From that point on, the identifier is sitting on OpenAI’s domain, built to leave chatgpt.com and show up elsewhere.

    What advertiser sites send back to OpenAI

    Any company that buys ChatGPT ads can install OpenAI’s measurement pixel on its own site. On a phone that already had __obi in the cookie jar, Buchodi saw the identifier go out on three kinds of requests:

    • The script load itself from bzrcdn.openai.com/sdk/oaiq.min.js
    • Conversion events posted to bzr.openai.com/v1/sdk/events
    • The SDK’s so-called no-credentials path, which still carried the cookie

    The script load is the ugly part. The browser attaches cookies to that request before any of OpenAI’s code even runs. Just loading the tag discloses the identifier.

    On the same advertiser pages, every other OpenAI cookie got blocked. Session cookies failed on domain mismatch. Cookies like oai-did were SameSite=Lax. __obi was the only OpenAI identifier configured with SameSite=None.

    What else rides along with the cookie

    The SDK does not just ping home. It also collects identity from the advertiser’s page. OpenAI labels four sources. One is values the advertiser passes on purpose. The other three are values the SDK scrapes from form fields, rendered page text, and the tag-manager bus.

    In the observed traffic, scraped identity beat advertiser-supplied identity 685 events to 255. The tag-manager bus was the largest source of email. The SDK replaces window.dataLayer.push with its own function, also reads adobeDataLayer, and finds renamed Google Tag Manager layers by parsing the gtm.js script tag.

    Here is what got sent:

    • Email, phone, first name, and last name were SHA-256 hashed before transmission
    • Country, region, city, and postal code went in the clear
    • Postal code was the most-harvested form field, 100 events across 28 sites
    • URLs were cut down to origin plus path. None of 23,929 observed events carried a query string
    • Paths that still reached the collector included a medical condition, a debt-solutions funnel, and a litigation intake form

    Automatic matching was on for 638 of 881 pixels with a known setting, including every credit and lending advertiser in the set. OpenAI controls that from Ads Manager. There is a denylist for passwords, one-time codes, card numbers, SSN, date of birth, medical history, diagnosis, and court fields. That is a list. It is not the same thing as leaving the rest of your life alone.

    It follows you even when you are logged out

    On Buchodi’s device, one __obi value went to OpenAI from 12 commercial websites under 13 distinct pixel IDs. The list included Chewy, Wayfair, ThriftBooks, Eventbrite, HelloFresh, Coursera, and SeatGeek. Every request came back 202, which means the collector accepted it.

    In the broader traffic, 12 of 30 distinct __obi values showed up under more than one advertiser. One showed up under ten.

    It is not just logged-in users. Across 932 decoded sync tokens, 736 carried subject_type account_user and 196 carried anonymous. The anonymous subject was as stable as the account subject: one per device, lasting at least 27 days.

    OpenAI calls this analytics, not marketing

    OpenAI’s cookie policy lists __obi under Analytics cookies, one year, on chatgpt.com and openai.com. It is the only entry in that section. The policy says analytics cookies help OpenAI understand how its services perform and are used.

    OpenAI splits consent into two buckets, oai_consent_analytics and oai_consent_marketing. Every sync token Buchodi decoded carried consent_decision: analytics_allowed. If you allow analytics and refuse marketing, you still get this cookie. That is not some blogger’s theory. That is how they labeled it.

    Buchodi sent the mechanism and two questions to press@openai.com and privacy@openai.com on 14 September: why is __obi classified as an analytics cookie, and does a user who grants analytics and refuses marketing still receive it. OpenAI Support acknowledged the inquiry, said the observations would be shared internally for review, and answered neither question.

    What this actually means for regular people

    This is standard adtech plumbing. Meta built the structural equivalent years ago: a logged-in account, third-party cookies on pixel fires, off-site conversions resolved to a profile. Buchodi is clear about that. What has no precedent is running it on an AI chat product.

    People tell ChatGPT things they would not put on Facebook. Health questions. Money problems. Legal messes. Then they go shop, read, or fill out a form on some other site that bought ChatGPT ads, and the same company that heard the chat now gets a signal from that page, tied to the same identity.

    Advertisers cannot even see this. __obi belongs to a domain their scripts cannot read. They installed a conversion pixel and have no way to know their visitors are being resolved to a ChatGPT identity. There is a different cookie, __obref, that lives on the advertiser’s own domain. Each site gets a different value. Of 2,860 values observed, 2,828 appeared under exactly one advertiser. That one is not the cross-site glue. __obi is.

    A few limits matter so nobody oversells this. It was observed on Chrome for Android. Safari’s Intelligent Tracking Prevention blocks all third-party cookies, and Chrome on iOS runs on WebKit, so this mechanism does not operate on any iOS browser. Desktop Chrome was untested. Roughly one ChatGPT session in five produced a sync token. ChatGPT’s mobile web client serves ads without syncing at all. And a 202 from the collector means the event was accepted with the cookie attached. That OpenAI resolves it to the account on the server follows from the design. Buchodi did not watch the join happen.

    What you can do about it right now

    • If you use ChatGPT, treat analytics consent as the switch for this cookie, not the marketing toggle
    • On Android Chrome, third-party cookies are how this travels. Blocking them, or using a browser that already does, cuts the cross-site path
    • iPhone and iPad users are already outside this particular mechanism because of WebKit
    • Do not assume logging out of ChatGPT saves you. The anonymous identifier was stable for weeks
    • If you run a small business site, understand that dropping an OpenAI ads pixel is not just measuring your own conversions. You are helping OpenAI stitch visitors to ChatGPT identities you cannot see

    I have spent 30 years watching companies tell regular people a cookie is just analytics while it does something else. This one is built to cross sites, lasts a year, and sits on an AI product people treat like a private conversation. That is not a feature for you. That is a system working exactly as designed, for them.

    Primary Source: https://www.buchodi.com/chatgpt-now-knows-what-you-do-on-other-websites-via-ad-collector/

  • Fake LastPass GitHub Downloads Pushed Rapuncel and Went After Your Passwords

    Fake LastPass GitHub Downloads Pushed Rapuncel and Went After Your Passwords

    In September 2026, a malware campaign impersonated LastPass on GitHub and tricked people into installing an information stealer called Rapuncel. That malware is built to steal passwords. If you went looking for a LastPass download and grabbed it from a GitHub page, you may have installed a thief instead of a vault.

    This is not some exotic lab stunt. It is a bait-and-switch aimed at regular people who just wanted their logins in one safe place.

    Fake LastPass pages on GitHub

    LastPass is a password manager. People use it for banks, email, shopping, and work. Attackers stood up GitHub downloads that pretended to be LastPass. The files were not LastPass. They were malware.

    GitHub is where developers host code. It is not the official place to get a consumer password app. That gap is the whole scam. You search, you see a repo with the right name, you hit download, and the system fails you in one click.

    Rapuncel steals what you were trying to protect

    Rapuncel is an information stealer. In plain talk, it reaches into a computer and pulls passwords and other login data. Pair that with a fake LastPass installer and you get a nasty result. Someone looking for a vault instead ran a tool built to empty it.

    I have watched systems take advantage of regular folks for more than 30 years. This one is blunt. You asked for a lock. They handed you a pick.

    Why this trap works on regular people

    People who live in GitHub check the owner, the history, and the code. Most folks do not. They want a password manager that works. They search, they click, they install. That is who this campaign is built for.

    It pisses me off because the people who would benefit most from a password manager are the same people least equipped to spot a fake repo. A small business owner. A parent trying to stop using one password everywhere. They should not need a security degree to download a password app.

    What to do before you install anything

    Keep this simple. Treat GitHub like a workshop, not a store, when the product is a password manager. I would tell my own family the same thing.

    • Only install LastPass from the official LastPass website or the official app store for your phone or browser.
    • Do not download a password manager from a GitHub repository, a mirror site, or a random search result.
    • If you already installed LastPass from GitHub, assume that machine is compromised. Change important passwords from a different, clean device.
    • Turn on two-factor authentication on email and banking so a stolen password is not enough by itself.
    • If you are not sure a LastPass download is real, do not install it. Go to the vendor’s own site and start over.

    Your passwords are the keys to your life. Do not pick them up off the side of the road just because the sign said LastPass.

    Primary Source: https://cyberinsider.com/fake-lastpass-downloads-on-github-pushed-password-stealing-malware/

  • DOJ Wins Again: Virginia Court Orders Google to Open Ad Tech to Rivals

    DOJ Wins Again: Virginia Court Orders Google to Open Ad Tech to Rivals

    The Justice Department’s Antitrust Division won substantial relief in its monopolization case against Google in advertising technology markets. In United States et al. v. Google LLC, the U.S. District Court for the Eastern District of Virginia ordered significant behavioral relief on September 16, 2026, including a requirement that Google closely integrate its products with products offered by rivals.

    That is the story. The government went after Google for locking down the plumbing of online ads. A federal court in Virginia told Google it has to open that plumbing to competitors instead of running the whole machine itself.

    What the Virginia court actually ordered

    This was not a press conference and it was not a fine they can write off as a cost of doing business. The court ordered behavioral relief. In plain English, Google has to change how its ad tech products work.

    The order includes requiring close integration between Google’s products and products offered by rivals. If you sell ads, buy ads, or run a website, that is the piece that matters. Google cannot keep its tools walled off while everyone else is stuck using whatever Google decides to allow.

    The case is United States et al. v. Google LLC. The “et al.” means other plaintiffs stood with the government. The Justice Department called this a win of substantial relief, and it said the department won again. This is not Google’s first time through this fight.

    Why Google’s ad tech grip matters to regular people

    Most folks never hear the words advertising technology unless they run a business or a website. You still pay for it. Every time a small shop buys an online ad, or a local site tries to sell space on its pages, that money runs through a stack of software that matches buyers and sellers in a fraction of a second.

    When one company owns too much of that stack, the people who actually make things and write things do not get a fair shake. Publishers get squeezed. Advertisers overpay. You see worse ads, and the websites you like have less money to keep the lights on. That is how a monopoly in this market shows up in real life.

    It pisses me off when a system this big takes advantage of regular people who have no seat at the table. The court stepping in is the law doing what it is supposed to do. Keep a market from turning into a company town.

    Behavioral relief is not a breakup

    Read the words. The court ordered significant behavioral relief. That is a change in conduct. It is not the same thing as splitting Google into pieces. Do not let anyone tell you this order is a breakup when the relief is about how products have to work with rivals.

    Close integration with rival products is a big deal if it is real and if somebody enforces it. It means Google’s ad tools have to talk to the other guy’s tools. Choice only exists if the pipes actually connect. If they do not, competition is a slogan, and the same company still runs the auction, the scoreboard, and the ticket window.

    What this means if you buy ads or run a site

    You do not need to become an antitrust lawyer. You do need to pay attention if online ads are part of how you make a living.

    • If you advertise a business online, more rivalry in ad tech is supposed to mean better prices and more options over time. Watch your costs. Do not assume Google is your only door.
    • If you run a website or a publication that sells ads, rival tools that actually plug into Google’s stack could give you more leverage. Ask whether anything changed in the contracts and the software, not just in the headlines.
    • If you are a regular user, you will not see a new button tomorrow. The point is whether the market behind the ads gets less lopsided.
    • Enforcement is the whole game. An order on paper means nothing if nobody checks whether Google really opened the door.

    I have spent more than 30 years helping small businesses cut through this kind of noise. The pattern is always the same. Big systems work great for the people who own them. Everybody else gets the leftover. When a court orders a company like Google to integrate with rivals, the question is simple. Does the system start working for more than one player, or does it stay a closed shop with extra paperwork?

    Keep your own data. Keep your options. Do not sign anything just because a monopoly lost in court and a sales rep suddenly sounds friendly. The order is from September 16, 2026. The work of making it real starts after that.

    Primary Source: https://www.justice.gov/opa/pr/department-justice-again-wins-substantial-relief-against-google

  • Telegram Desktop Flaw Let Bot Buttons Steal Your Chats From HTML Exports

    Telegram Desktop Flaw Let Bot Buttons Steal Your Chats From HTML Exports

    A high-severity flaw in Telegram Desktop allowed attackers to steal chat messages through poisoned HTML exports. Malicious JavaScript hidden inside bot-created buttons could grab conversation content when you exported that chat as HTML and opened the file.

    That is a nasty little chain. It starts with a bot sitting in a chat you already have, and it finishes with your private messages leaving your computer.

    How a Bot Button Turns Into a Chat Thief

    Telegram bots send buttons all day long. Confirm an order. Open a link. Vote in a poll. Those buttons are supposed to be harmless controls. The problem is they could carry malicious JavaScript.

    Inside the live Telegram Desktop app, that kind of code is supposed to stay locked down. The break happened on export. When you saved a conversation as HTML, Telegram wrote out a real web page. Open that page in a browser and the script in the bot button can run the same way a script runs on any website.

    Once it runs, it can read the chat sitting in that file and send it off the machine. You thought you were making a backup. You handed over a copy.

    Why the HTML Export Is the Weak Link

    People export chats for normal reasons. A small business keeps customer threads. A family saves a group before someone leaves. A shop owner wants a paper trail. HTML looks like the real conversation, which is why Telegram offers it.

    HTML is not a document. It is a webpage. Webpages run code. JSON does not. A plain text dump does not. If a hostile bot planted JavaScript in a button before you hit export, your browser will treat that poison as part of the page.

    You do not need to be famous for this to matter. You just need to export a chat that included the wrong bot, then open the file.

    What Regular Telegram Users Should Do

    This one gets under my skin because it punishes the responsible move. You saved a copy. The copy bit you. Big platforms keep adding features and then forget that regular people will use those features exactly as advertised.

    If you use Telegram for work, a community group, or family, treat HTML exports like live webpages. Do not email them. Do not toss them in a shared folder. Do not open old ones you cannot account for.

    • Update Telegram Desktop and leave auto-update on. These holes live longest on old installs.
    • Be picky about bots. If you did not go looking for that bot, do not keep it in the chat.
    • Do not open HTML chat exports from other people. Treat those files like a random attachment.
    • If you need a backup, think twice before choosing HTML. A format that cannot run code is the safer copy.
    • If you already exported chats that had unknown bots in them, do not reopen those HTML files until you know what you are looking at.

    Keep the System Straight in Your Head

    Bots are not your friends. Export is not a harmless Save As. A file that ends in .html is a webpage, not a printout. Hold those three facts and this whole class of problem gets a lot less dangerous.

    Telegram feels private and easy. Easy is where corners get cut. When chat furniture like a bot button can ride along into a file your browser will execute, regular people pay for it. That is the system failing, not you being careless.

    Primary Source: https://cybersecuritynews.com/telegram-desktop-flaw/

  • Telegram Desktop Flaw Let Attackers Steal Your Chats Through Poisoned HTML Exports

    Telegram Desktop Flaw Let Attackers Steal Your Chats Through Poisoned HTML Exports

    A high-severity flaw in Telegram Desktop let attackers steal chat messages. Malicious JavaScript hidden in bot-created buttons could run when a conversation was exported as HTML and then opened. That is not some lab curiosity. It is a poisoned copy of your own chat.

    Telegram Desktop lets you export a conversation as an HTML file so you can read it in a browser. People do that for records, handoffs, or a backup they actually control. The hole sat in how those exports handled buttons that bots dropped into the chat.

    Bot Buttons Should Not Execute in Your Backup

    Bots on Telegram can send messages with buttons. Inside the app, those buttons are supposed to stay inside Telegram. In the HTML export, malicious JavaScript stuffed into those buttons could ride along in the file.

    Open that export in a browser and the script can run. Once it runs, it can steal the chat content sitting in that same file. You did not have to click a shady link that day. The trap was already in the conversation. The trigger was a normal backup.

    This Is How Regular People Get Burned

    Small shops use Telegram. Families use it. People who want a little distance from the usual big platforms use it. An HTML export is often the only local copy of a long thread.

    If an attacker poisons that file, they are not skimming one joke. They can walk off with the whole conversation. Customer names. Family plans. Whatever you treated as private because it lived in a chat.

    It pisses me off when a feature that is supposed to give you your own data turns into a delivery system for somebody else. You should not need a computer science degree to save your own messages.

    What You Should Actually Do

    • Do not open Telegram HTML exports in a regular web browser if the chat included bots you do not fully trust.
    • Treat an HTML export like it might contain live code, not like a harmless printout.
    • Keep Telegram Desktop updated. Old desktop clients are where holes like this hang around.
    • Be stingy with bots in chats that hold anything you care about. An unknown bot is an unknown person with a keyboard.
    • If a conversation matters, do not assume export equals a safe archive.

    Control over your own data only works if the copy you make does not execute someone else’s script. That is the whole point of having a backup in the first place.

    Primary Source: https://cybersecuritynews.com/telegram-desktop-flaw/