The IT Nerd

OODA Loop reports today that “Hackers Have Uploaded Thousands Of Malicious Files To Hugging Face Repository” based on input from Protect AI.  

The OODA Loop story reads in part: “The old Trojan horse computer viruses that tried to sneak malicious code onto your system have evolved for the AI era,” said Ian Swanson, Protect AI’s CEO and founder.

  “The Seattle, Washington-based startup found over 3,000 malicious files when it began scanning Hugging Face earlier this year. Some of these bad actors are even setting up fake Hugging Face profiles to pose as Meta or other technology companies to lure downloads from the unwary, according to Swanson. A scan of Hugging Face uncovered a number of fake accounts posing as companies like Facebook, Visa, SpaceX and Swedish telecoms giant Ericsson. One model, which falsely claimed to be from the genomics testing startup 23AndMe, had been downloaded thousands of times before it was spotted…”

Mali Gorantla, Chief Scientist at AppSOC had this to say:

  “It should surprise no one that Hugging Face has become a magnet for malware and bad actors. In the last year, the number of AI models available on Hugging Face has tripled, now topping 1 million. Data scientists and AI developers love experimenting with this vast amount of open-source data to build and train new AI applications. The problem is that most security teams have little visibility into what models or datasets have been downloaded or where they exist. I can’t think of a more obvious place to embed malware, infiltrate corporate defenses, and hide your tracks.”

Security teams need to change their tactics so that they have visibility and are able to uncover this sort of thing. Because this is clearly the next “big thing” that threat actors are engaged in.

The Internet Archive has experienced yet another breach, this time involving their Zendesk email support platform with 800+ support tickets, despite warnings about stolen GitLab authentication tokens by threat actors.

“It’s dispiriting to see that even after being made aware of the breach weeks ago, IA has still not done the due diligence of rotating many of the API keys that were exposed in their gitlab secrets,” reads an email from the threat actor.

Since Saturday night, BleepingComputer reported receiving multiple emails from individuals who got replies to old Internet Archive removal requests, alerting them to the breach caused by the organization’s failure to properly rotate their stolen authentication tokens.

Recipient of these emails told BleepingComputer that they had to upload personal identification when requesting a removal of a page from the Wayback Machine.

“Whether you were trying to ask a general question, or requesting the removal of your site from the Wayback Machine your data is now in the hands of some random guy. If not me, it’d be someone else,” the threat actor’s email continues.

Steve Hahn, EVP Sales US, BullWall had this to say:

  “Multiple successive attacks is unfortunately the norm, not the exception. When a threat actor has a successful attack they have typically spent months in the environment undetected. They have worked for long term persistence. Setting up dozens or hundreds of back up accounts and credentials, running scripts to cover their tracks, set up fresh, unprotected VMs, done vulnerability scans, laid second wave traps, such as embedding malicious macros into internal documents that will launch a whole new attack. The latter is quite crafty. We all know we shouldn’t “enable macros” on any file we get from an untrusted source, but when it’s on an internal share and it’s a document you use regularly, you have no hesitancy to hit the “enable macros” button. In terms of how often a company is hit in successive attacks, I’ve seen numbers as high as 78% and that does ring true to my personal observations.”

It’s bad enough that this site got pwned. But to get pwned three times is insane. Hopefully the Internet Archive takes steps to make sure that there is not a fourth time as this is pretty embarrassing.