Do we control the technology we use?

Social networks are a great vehicle for sharing. Each user contributes by sharing their own experiences. Our smartphone has more information about us than our parents or our partner do. This allows us to create relationships and improve existing ones, enabling us to get to know others better.

One of the many drawbacks is that all the data we put on social networks is accumulated by companies that become its owners. These companies have a view of the world that is more comprehensive than anyone has ever had before. Knowing people’s tastes, orientations, thoughts, and moods allows them to be monitored, but above all, it enables them to be solicited, guided, and ultimately led.

This happens with social networks and with the lives of the people who participate in them. I won’t comment on who profits from this; there are journalists who deal with these things on a daily basis and are certainly more reliable than I am.

However, this mechanism does not work well with companies. Of course, by looking at a company’s website, we can see what it does. By looking at LinkedIn, we can understand who works there and possibly what they do and how they do it. It always remains at a high level; from the outside, we can peek in, but not much more. Typically, unless there is a leak, we know what a company wants us to know.

So how could someone from the outside really understand how a company works? How could someone understand the business processes, methodologies, and technologies used in a particular company?

The solution could be to create an assistant based on artificial intelligence!

Once this is done, it is advertised as the future, with speculation that it will allow anyone to do anything, and surely someone will find it convenient to share everything they would not even confide to their mother in a chatbot (about which they know nothing).

Because in exchange for their secrets, the chatbot gives users something they could not otherwise have (except through hard work and perhaps even study): the feeling of being autonomous. Even in contexts unknown to the user. No need to ask, no need for professionals, no need for culture and knowledge. Artificial intelligence does everything. Just pay and share. So you pay twice. Brilliant!

You don’t know how to use Photoshop but you want an image? There’s GenAI! You don’t know how to write code but you want to make an app? Long live Vibe Coding!

In the end, it doesn’t matter that you had to share your secrets with a stranger to get a solution, because the user thinks they did it all themselves in the silence of their office. Them and their computer.

It doesn’t matter where the information ended up because they got an answer in return. They did it all themselves. A kind of masturbation, live on TV.

A doubt about eBay

After weeks of eBay stressing me because I don’t have a payment method associated, I decided to do as they ask and associate a payment method. I thought a credit card would be fine, but no, they want a bank account. I start the procedure: they ask me for the bank I use, a popup from my bank opens asking me for credentials, I enter them and here, rightly, the bank asks me to agree to provide authorizations to eBay. Here I notice two things that I don’t understand: first, the authorizations that eBay is asking me to grant are related, not only to the checking account, but also to a credit card. In addition to this, which already seems excessive to me, the list of authorizations includes: coordinates, balance and list of transactions.

I could also find a justification for these requests, but more easily I see a desire to mind my own business.

How to use a Telegram token

In the following article I talk about some approaches to mitigate the damage from a criminal attack. I am not sure if what I am talking about is legal in all countries, so do your research first. Or use a VPS in Russia 😀

Telegram is one of the main tools used by criminals to receive stolen information through various types of attacks, including of course phishing sites.

Below I outline some simple actions we can take to mitigate Telegram bot-based attacks. But first let’s see how to use a token.

For this post I will use a token identified in these hours within a kit distributed on different domains:

  • psyclum.org
  • lwonsio.org
  • peyssc.org
  • katjrig.org
  • katjriv.org
  • sorajfm.org
  • niondin.org

Analyzing the deployment methods of the kit, we can see how the criminal registered the various domains and only after several days installed the kit. This in an attempt to avoid being intercepted. Fortunately, Matrix also manages these situations 🙂

Let’s move on to the kit, as always it’s a piece of junk written in PHP. Inside we find the Telegram bot token to which stolen credentials are sent to unwary users.

Now let’s move on to the main topic of this article: we have a token, what do we do with it? First, let’s study. Here is the documentation on how to use bots: https://core.telegram.org/bots/api

Using VS Code and the amazing “REST Client” extension we can query bots and use it to mitigate criminal action.
First, using the “getMe” method I check if the token is active. The 200 return code and the Json in response indicate that the token is active and therefore ready to receive information from clients, including the phishing site itself.

When the unwary user enters their credentials into the phishing site, it sends the information to the Telegram bot. Two things can happen here: if there is a registered webhook, the information is sent to the resource indicated as the webhook, otherwise the information is kept in the bot’s message queue to be consumed via the “getUpdates” method.

Using the “getWebhookInfo” method we check the bot status and the possible presence of a webhook. In this case the webhook is not registered and therefore we must use the “getUpdates” method to download any messages present in the queue.

By calling the “getUpdates” method we see that there are no messages in the queue. If there were any, our call would have returned them to us by deleting them from the queue.

As mentioned, we have several ways to handle this situation. The first is to register a resource of our own as a webhook that will then receive the stolen information, allowing us to act promptly. For example, you could implement an alert system for the user to warn him that his credentials have been stolen and perhaps even deactivate his account and credit cards. As an alternative to registering your own webhook, you can implement a polling mechanism to download the stolen information, removing it from the criminal’s reach, and use it to mitigate the threat: as mentioned above, you can block the account, warn the user, and explain to them that they need to be careful.

Another approach is token deletion. I personally don’t like this approach because it prevents us from intercepting compromised accounts and therefore deprives us of the ability to act on users.

Phishing using SVG

Today I found an email with an SVG attachment in my secondary email account. The account is Office 365 and the email was not in spam, this is not good, especially considering how much 365 costs! Anyway, back to the email, it was clearly suspicious also considering the attachment, an svg file with the name “Your-to-do-List.svg” supposedly sent to me by itself.

I immediately downloaded the file and opened it with Notepad++ (all this without the antivirus having anything to say about it…), inside there was a script that through a simple algorithm made a redirect to a phishing site with the aim of stealing the credentials of the same Office 365 account.

The target URL is: hxxps://bzzrr[.]yiunwox[.]es/!7ijf9v9ehpNFEKt/$

When I tried to forward the message to my Gmail account, Google’s mail server correctly returns an error 🙂

As always, avoid downloading and opening attachments if you are not sure of the source.

Artea and ING phishing kit

Yesterday I came back from a weekend in Latvia and now I have come across an attack on Artea, a Lithuanian bank… The Baltics are calling me 🙂

Matrix has identified a series of patterns that have allowed them to quickly identify the various domains involved:

  • banklithuanial.net
  • lithuaniabankasl.net
  • arteasite-login.net
  • arteabankslogin.net
  • italyingbank.net

One of the sites used for the attack on Artea hosts an attack on ING Italia based on a variant of the same kit.

The kit is simple but decently done.

The comments are in Russian and the Telegram tokens used to transmit the stolen credentials are these:

  • 7621096866:AAFW8cGs93gcFPbpFmyK-mMJKTYYLp4ENwg
  • 7764061568:AAGVrwI8IukR8kqgjn_mEqywmHHtE3RJywE

Correios phishing kit

In the last few weeks I have noticed that attacks on Correios are constant. I have studied the matter a bit: it is the Brazilian state company that manages shipments and payments related to them.

The scam is always the same, attackers write to users saying that a shipment is blocked and that a small payment is needed to unblock it. What is interesting about this attack is that the user by entering the CPF (Cadastro de Pessoas Físicas) gets his data displayed on the page and this certainly appears to the user as something reliable.

The domain used for the attack is cpf-pendente[.]co[.]ua and it was registered few hours ago.

The first curiosity is that the archive is a RAR instead of a ZIP. Someone hoped to make life difficult for Matrix but I love writing code and this allows Matrix to open different types of archives, including obviously RAR 🙂

Aside from the type of archive, my curiosity about this kit was given by the functionality of extracting user data starting from the CPF. I do not know the Brazilian law but that from a simple number you can extract personal data seems unsafe to me.

The kit uses an external service to resolve the client data. This service I assume is part of the attack and probably uses another service in turn, perhaps legitimate or a leak from who knows where. This service is available at the domain anonimobusca[.]online. It was registered one week ago.

To obtain the user information, the phishing kit, uses a specific resource exposed by the second domain

Another aspect is related to how payments are collected. The kit does not in fact collect itself he payment data. The user makes the payment using systems that manage transactions. The services used to perform the transactions are:

  • paguesafe[.]com
  • atlaspagamentos[.]com

The URLs used are:

  • hxxps://checkout[.]paguesafe[.]io/pl/mn7xXvjg2N
  • hxxps://checkout[.]atlaspagamentos[.]com/pl/RQ9Kz7XTf8

Both services have been active for a short time, they use cheap providers for domain management and have free certificates. In short, I would say that nothing seems to give trust to these services.

To this I would like to add that the two graphical interfaces are practically the same.

Being a hacker

Nothing annoys me more than hearing the term “hacker” used as a synonym for “criminal”. I can actually accept it from people who do something else in life. It really bothers me when this abuse comes from people who say they are cyber security professionals.

I understand that in recent years everyone has become cyber security experts. Those who were previously blockchain experts, before big data experts and even before one of the many trends that have plagued IT. In many cases, the place of many of these would have been a nice industrial fryer for chips in a fast food restaurant.

“I know that I know nothing”, having quoted Socrates, I can admit that I am not the right person to explain what a hacker is, so I report the Manifesto.

If you don’t work in IT and you don’t know it, it’s not serious.
If you work in IT and you don’t know it, it’s serious.
If you work in cyber security and you don’t know it, for me mayonnaise and ketchup 😉

Being a hacker means facing life with passion and living the passion without limits. This does not mean harming others. For this reason a hacker is not a criminal. He can be if he harms others, but in doing so he becomes a criminal and must be considered as such.

Another thing: let’s be serious, “ethical hacker” is funny, it smacks of an elderly “script kiddie”!