Archive for ExpressVPN

Chatbot logs and audio exposed in data breach at major department store chain

Posted in Commentary with tags on March 17, 2026 by itnerd

Cybersecurity researcher Jeremiah Fowler recently discovered 3 separate databases that were neither password-protected nor encrypted and contained a total of 3.7 million chat log transcripts, audio recordings, and text transcriptions of phone calls exposing Sears Home Services.

The publicly exposed databases totaled over 4TB and contained:

  • 2,116,011 txt files that exposed names, phone numbers, physical addresses, and user-submitted personally identifiable information (PII).
  • 207,381 xlsx files and audio recordings totaling 415.2GB.
  • 1,442,577 audio recordings of customers and their text transcripts totaling 3.9TB.

Jeremiah’s detailed findings are published on the ExpressVPN blog here: https://www.expressvpn.com/blog/searshomeservices-data-exposed/.

2015 vs. 2025: How password habits have evolved over the past 10 years

Posted in Commentary with tags on March 6, 2026 by itnerd

ExpressVPN has published an article on the evolution of password security over the past 10 years. Cybersecurity researcher Jeremiah Fowler has published an analysis of part of the data from the recent 149 million credentials leak on the ExpressVPN blog, comparing current password habits with those from a decade ago.

During this research, Jeremiah noted some interesting and concerning findings:

  • Only 15% of the passwords from 2025 could be classified as complex.
  • 85% of current passwords typically contain known patterns from prior breaches or password-guessing models.
  • It’s still common for people to reuse passwords across multiple accounts.

Jeremiah published his detailed report on the ExpressVPN blog here: https://www.expressvpn.com/blog/password-security-2015-vs-2025/

149M harvested credentials exposed in data breach 

Posted in Commentary with tags on January 23, 2026 by itnerd

Cybersecurity researcher Jeremiah Fowler recently discovered a non-password-protected database containing over 149 million unique credentials. These records were collected from victims of malware worldwide and include everything from social media and streaming services to sensitive financial logins.

In a few words, the publicly accessible database:

  • Exposed 149,404,754 unique logins and passwords (96GB of raw data);
  • Revealed user credentials for major platforms (including Facebook, Instagram, TikTok, X, dating sites, and OnlyFans, affecting both creators and customers);
  • Included high-risk financial credentials (such as crypto wallets, trading services, and banking logins).

Because this data was likely collected by malicious third parties, there is a heightened risk of widespread credential-stuffing attacks, identity theft, and financial fraud. 

Jeremiah published his detailed findings on the ExpressVPN blog here: https://www.expressvpn.com/blog/149m-infostealer-data-exposed/

UPDATE: I have commentary on this starting with Paul Bischoff, Consumer Privacy Advocate at Comparitech

“The data is a gold mine for cybercriminals launching credential stuffing attacks. Cybercriminals can use stolen username and password combinations to log into a wide array of accounts under the assumption that many people use the same password across multiple accounts. This process is automated, so a hacker can attempt to use a single set of credentials across dozens or even hundreds of accounts in a matter of seconds.

This data exposure highlights the importance of setting unique passwords and using two-factor authentication when available. If you don’t reuse passwords, then you are immune to credential stuffing attacks. Even if a cybercriminal tries to log into your account with the correct password. two-factor authentication will prevent them from doing so in the vast majority of attacks.”

Chris Hauk, Consumer Privacy Champion at Pixel Privacy:

“The report indicates the harvested login credentials were the results of “Keylogger” and other types of “infostealer” malware underscores the need for computer users to run Antivirus and ant-malware protection on their machines. Whether they use Windows or macOS, there are risks to not keeping your machine safe by running security apps in the background. 

The exposure of such a huge number of credentials poses a significant risk to users that are not aware of the breach and to what extent they are exposed. While it may be too soon to have this information included in the “HaveIBeenPwned” (https://haveibeenpwned.com/) website’s extensive database, I still strongly recommend that users visit the site and enter their email address(es) to determine whether their information has been exposed in previous data breaches. I also recommend that they take advantage of the website’s option to notify them when their email address was exposed in future data breaches.

Last but not least, everyone should use a password manager. In addition to keeping track of login information for multiple sites, password managers often offer warnings about password reuse or if a login has been exposed in a breach. This makes it easy to guard against password reuse, and to update passwords when they need to be changed.”

Data breach affecting AI image generator, exposing sensitive images 

Posted in Commentary with tags on December 5, 2025 by itnerd

Cybersecurity researcher Jeremiah Fowler recently discovered a non-password-protected database containing over one million sensitive records belonging to Magic Edit, a popular AI image generator tool developed by BoostInsider Inc.

In a few words, the publicly accessible database:

  • exposed 1,099,985 images and video files;
  • included face-swapped images on AI-generated bodies converted into sexually explicit images;
  • contained unaltered images of real individuals, possibly uploaded as references and presumably without those individuals’ knowledge or consent.

Jeremiah published his detailed findings on the ExpressVPN blog which can be found here: https://www.expressvpn.com/blog/magicedit-data-exposed/