The new AI model Grok-4 was hacked just 48 hours after its release

In an event that has surprised the tech community, the new artificial intelligence model Grok-4 was released just 48 hours after its launch. This rapid vulnerability reveals not only weaknesses in software security but also the growing ability of hackers to find and exploit flaws in sophisticated systems. Record time The methods used to carry out the jailbreak of Grok-4 have been identified as Echo Chamber and Crescendo. The first, Echo Chamber, focuses on creating a controlled environment where the system’s responses can be manipulated, facilitating unauthorized access. Crescendo, […]

In an event that has surprised the tech community, the new artificial intelligence model Grok-4 was released just 48 hours after its launch. This rapid vulnerability reveals not only weaknesses in software security but also the growing ability of hackers to find and exploit flaws in sophisticated systems.

Record Time

The methods used to carry out the jailbreak of Grok-4 have been identified as Echo Chamber and Crescendo. The first, Echo Chamber, focuses on creating a controlled environment where the system’s responses can be manipulated, facilitating unauthorized access. Crescendo, on the other hand, represents a more aggressive tactic of information accumulation and exploitation of weaknesses in the software architecture. Both techniques have been employed in the past in attacks on other artificial intelligence systems, suggesting a pattern that cybersecurity researchers will need to analyze swiftly.

The launch of Grok-4, which took place recently, had been anticipated as a significant advancement in artificial intelligence technology, promising more accurate and efficient algorithms. However, its rapid exposure to attacks raises concerns about security in the development of such products. The development community and cybersecurity experts are now on alert, seeking ways to protect future iterations of artificial intelligence and to prevent similar incidents.

Rumors in the sector suggest that Grok-4 may not be the only system with similar vulnerabilities, indicating that this type of attack could be more common than previously thought. Meanwhile, developers are expected to take immediate action to address these security issues, thus protecting both their technology and end users.

Be careful with the companies that call you on the phone! It could be a scam

You can no longer trust even your lifelong brands: several cybersecurity researchers have warned about an increase in phishing campaigns that use impersonation of recognized brands to deceive victims, leading them to call numbers controlled by malicious actors. This technique, known as telephone-oriented attack delivery (TOAD), relies on victims communicating with attackers under the premise of resolving a problem or confirming a transaction. “Tell me?” Well, it’s a scam According to a report from Cisco Talos, among the most impersonated brands are Microsoft and […]

You can no longer trust even your long-time brands: several cybersecurity researchers have warned of an increase in phishing campaigns that use impersonation of recognized brands to deceive victims, leading them to call numbers controlled by malicious actors. This technique, known as telephone-oriented attack (TOAD), relies on victims communicating with attackers under the premise of resolving a problem or confirming a transaction.

Tell me? Well, it’s a scam

According to a report by Cisco Talos, among the most imitated brands are Microsoft and Docusign. Phishing emails that include legitimate PDF files, such as those from Adobe, increase the effectiveness of these attacks. Criminals take advantage of a Microsoft 365 feature that allows them to send phishing emails without needing to compromise accounts, affecting more than 70 organizations.

These attacks, which often convey a sense of urgency, employ sophisticated social engineering tactics, mimicking real customer support workflows. Researchers point out that live interaction through phone calls allows attackers to manipulate victims’ emotions and responses, increasing the risk that they will disclose sensitive information or install malware on their devices.

The use of artificial intelligence-powered tools has facilitated the creation of large-scale phishing pages, and researchers have also warned about the possible use of language models to flood queries with malicious links. Additionally, attempts have been detected by malicious actors to influence the search results of legitimate sites, using techniques to integrate phishing links into compromised pages.

These practices not only threaten the security of organizations, but also erode user trust in brands, highlighting the need for effective solutions to detect and prevent brand impersonation in the cyber realm.

iPhones can be hacked just by receiving a text

This bug can allow hackers access to all of your personal data just by receiving an iMessage!

iPhone

We’ve all had close calls when it comes to getting our phones hacked.

Sometimes we click on a malicious link while browsing the internet. Sometimes we download a shady app that is sending data in the background. All of these things typically boil down to our own human error whether it be from lack of knowledge, or taking a chance on a shady app. However, it was recently discovered that an iPhone can be hacked simply by receiving an iMessage. This has brought more attention to a  relatively new type of hacking tool known as “interaction-less” bugs.

Should I be worried?

At the Black Hat Security Conference this week, a Google Project Zero researcher showed that there are “interaction-less” bugs in iMessage that hackers can exploit to gain control. Once hackers gain control, they will be able to access your personal data on your phone.

You can click here to see the researcher’s full slideshow presentation. 

The bugs were found by looking for flaws while reverse engineering iMessage. The researcher, Natalie Silvanovich, did not find any similar bugs in Android phones. However, throughout the years, she and her team have found similar bugs in apps like WhatsApp and FaceTime. 

What should I do?

The best thing you can do is to make sure you are updating your phone whenever you can. According to Silvanovich, Apple has already patched six of these bugs in updates. 

You should also run regular virus scans on your phone to make sure that you are virus-free. We highly recommend using Malwarebytes since it is both free and efficient.

Malwarebytes Download now ►
8

You can also block texts from unknown numbers so that you won’t hear from them again. To do this, follow these steps:

1. Go to the Messages app

2. Tap on the spam message you received to open it

3. Tap at the top of your phone until the “Info” button appears.” Tap “Info”

Click info

4. Tap on the number at the top of the screen

Click number

5. Scroll down to the bottom, and tap “Block this Caller”

Block this Caller

6. You should stop receiving texts from this number altogether

If you do receive a shady text from an unknown number, you should block the caller, and then run Malwarebytes. Once you’re done, check and make sure that your phone is up to date.

Wrapping up

Although there is virtually no chance that your up-to-date iPhone will be hacked through this method, the question now remains: “How did Apple leave this vulnerability in the first place?” If you’re spending hundreds of dollars on an iPhone, you should rest assured that it won’t be easy for a hacker to access your personal data. Just recognize that your security depends on your vigilence.

Tor anonymous browser targeted by Russian intelligence

Tor users beware: your web-browsing might have been compromised.

Tor

Hackers have discovered that a Russian Intelligence contractor was trying to breach Tor.

What is Tor?

In case you’ve never heard of it, Tor is a web browser that allows you to surf the web privately without fear of government surveillance or being tracked. It is very popular in Russia although it is available in other countries.

Is Tor a good browser?

Aside from this recent news, it’s okay, but there are better options that we will get into later.

How were they trying to do this?

When someone uses Tor, their internet service provider gets a ping that lets them know that it is being used. However, the ISP doesn’t get notified about what websites they were surfing. 

The hackers stole more than seven terabytes of data from SyTech, the Russian Intelligence contractor.

In that data, they discovered that SyTech planned to use a program called Nautilus-S to discover what website a Tor user was on before closing the window. 

Did it work?

Oddly enough, this breaching plan seemed like more trouble than it was worth. Even if you were a Tor user in Russia, there was a slim chance that you were susceptible to the breach, and an even slimmer chance that the info would be sent back to SyTech. 

According to a statement from a Tor spokesperson, even if a Tor user was affected by the breach, they were still kept anonymous. 

What should I do?

Unless you live in Russia and worry about government surveillance, you probably aren’t a Tor user. If you are, you might want to switch your browser of choice. 

If you haven’t been convinced by all the wonderful security features we’ve already written about, then hopefully this news has convinced you to finally switch to Firefox or Brave.