Apple Processor Flaw Lets Hackers Steal Encryption Keys
Researchers have identified a new vulnerability in Apple’s M-series processors for Macs. The vulnerability, named GoFetch, allows an attacker with a malicious app installed on a Mac to eventually defeat portions of its encryption. An attacker can cycle through many possibilities of what the encryption key might be, and the vulnerability will leak a subtle indicator when an attempt is partially correct. This lets the attacker eventually learn the whole key. There is no evidence yet that this attack has been used, and Apple has been notified, however the vulnerability is in the design of M-series hardware and cannot be patched with software. App developers are able to mitigate the vulnerability in their own apps by employing a fix known to slow down their apps.
The Bottom Line: This vulnerability cannot be patched on M1 or M2 devices, but M3 devices may see an update that would slightly impact performance in order to fix the vulnerability for those machines. Exploiting the vulnerability does require that the attacker have code-execution ability on your machine, so it is prevented by safe consumer electronics use. Only install apps from reputable developers.
Change Healthcare Hit with Major Ransomware Attack
Change Healthcare is the largest payment clearing company for healthcare providers in the USA. In late February, many of their servers were taken offline in response to a ransomware incident by the criminal hacking gang known as BlackCat. James Rundle from the Wall Street Journal prives a good high level overview. Disabling the servers at Change Healthcare made it impossible for many healthcare providers in the USA to receive payment for up to 98 percent of the services they render. Providers ranging from single office family doctors to whole hospitals to care homes across the US have lost out on income ranging from a few thousand to millions every day. Many have survived through emergency assistance, others through loans, others have closed.
The BlackCat group responsible for the attack operates a ransomware as a service model, where they create and maintain the software necessary to encrypt a victim’s computer in exchange for taking a percentage of each ransom paid by victims, but they rely on independent operators, known as affiliates, to hack victims and install the ransomware that will encrypt its drives, lock up the computers, and potentially steal its data, until the ransom is paid.
It appears that Change Healthcare paid the ransom to get their computers back online, however, due to some criminal backstabbing from BlackCat, the ransom money never made it to the affiliate who had control of the computers: BlackCat took it and ran.
As of March 28, Change Healthcare is still not fully recovered.
The Bottom Line: The final cost to the US healthcare system is not yet known, but the effects are far reaching. We don’t know what, if any, personal information was stolen from Change Healthcare and is now in the hands of BlackCat or their affiliates. There isn’t anything that we as individual consumers can do except push elected representatives for better oversight and diversification of the financial services industries. Be on the lookout for scams trying to capitalize on the confusion, such as unsolicited offers of alternate payment systems. Always be wary of unsolicited calls or texts, especially if they try to pressure you with urgency or threats. You can always hang up, then call them back at an officially listed number. Always monitor your bank and credit card statements as well as your credit score and consider freezing your credit.
How to Steal a Tesla with Your iPhone
We do a lot of how-to videos but here's one we didn't make. Security firm Mysk released an explainer video demonstrating how a loophole in Tesla’s security practice would let an attacker steal a Tesla. The attack works by setting up a free Wi-Fi network at a Tesla charging station, but branding it to look like the free Wi-Fi offered at Tesla dealerships. Then the scammer just waits for someone to connect. The malicious Wi-Fi will ask for the victim’s Tesla ID and password to log them in. Once armed with that credential, the scammer can log in to the victim’s Tesla account and add their own iPhone as a key to the car, then simply unlock the car and drive away. The victim never receives any notifications of a new login to their account, of a new key added to their car, or that the car has been unlocked or started.
The Bottom Line: Tesla doesn’t offer free Wi-Fi at their charging stations, only at their dealerships. If you see free Tesla branded Wi-Fi at a charging station, it’s probably a scam. It’s best to avoid free Wi-Fi in general, because this is just one kind of scam of dozens that can be facilitated through malicious use of free Wi-Fi networks. Using multi-factor authentication on your Tesla account could prevent this attack, especially with a hardware key.
Public Database Reveals 5 Million Text Messages a Day
An internet and cell phone service provider called XY International, which primarily services Asia and claims to handle about 5 million text messages per day, accidentally left an internal database open to the internet without a password, reports Zack Whittaker at Tech Crunch. The database contained details of all text messages transiting their systems including for example the sensitive password recovery links for people trying to create a new password for a google or microsoft account, as well as the MFA codes used to secure accounts. Plus all of the private communications of millions of users, available to be read, copied, stored, and analyzed by anyone with the link.
The Bottom Line: Even if you are not a customer of XY International, this supplies evidence of the problems inherent to using text messages as an identity verification tool. When setting up multi-factor authentication (MFA) for your important accounts, bear in mind that a text message isn’t as secure as other forms of MFA. This also demonstrates the risks of using any unencrypted messaging system, and it's why we recommend using encrypted messagers like iMessage or the secure app Signal.
X Update Leaks Your Location
The social networking platform X (formerly Twitter) activated an audio or video calling feature for all users of its apps as of the 4th of March. A call on X permits both participants to see each other’s real IP address, which can be linked with their physical location. By default, the X app allows calls to you from anyone you follow, though you can limit that or expand it in the app’s settings. Despite its many security and privacy failings, X continues to be used by activists and journalists around the world whose safety is contingent on their locations remaining private.
The Bottom Line: If you’re still using the X app, consider disabling the video calling feature in the settings area. Tap your profile picture > Settings and Support > Settings and privacy > Privacy and safety > Direct messages, toggle off Enable audio and video calling. In that same menu there is a feature called Enhanced call privacy that will route the audio and video through X’s infrastructure to avoid giving away your IP address, but I would recommend using a more secure app for your calls.
Interactions with Chat GPT Are not Private
Researchers have developed a simple method for a passive observer of network traffic to identify the content of all Chat GPT interactions with high accuracy, even when they’re encrypted. The same approach works on every large language model search engine except Google Bard. Read more at Ars Technica.
The Bottom Line: While this is a novel tactic, it is one of many similar methods. For the moment it is probably best to consider any questions that you ask of large language models to not be private, and to assume the entire interaction can be inferred, recovered, or intercepted by third parties and the model operators. Do not use large language model generative AI to answer questions on sensitive topics or to evaluate your secret information.