A recent security breach on a social network for AI agents has left thousands of users' email addresses and millions of API credentials exposed, raising serious concerns about the vulnerability of these platforms to hacking. Moltbook, a Reddit-like platform designed for AI agents to interact with one another, was found to have a critical security flaw in its JavaScript code that allowed anyone to access user accounts and private communications.
The founder of Moltbook, Matt Schlicht, had touted the site as "vibe-coded" by his own vision, rather than having written any code himself. This lack of human oversight is believed to be responsible for the security flaw. The vulnerability was discovered by researchers at Wiz, a security firm, who found that it could allow complete account impersonation and access to private communications.
The discovery serves as a cautionary tale about the security risks of AI-made platforms, highlighting the need for more stringent controls on AI-generated code. Rather than inherent flaws in companies' implementation of AI, the problem often lies with allowing AI to write their own code - resulting in numerous bugs.
On a separate note, Apple's Lockdown mode has proven itself as an effective safeguard against government hacking attempts, including a recent FBI raid on a Washington Post reporter's home and search of her devices. The Lockdown mode feature prevents connection to peripherals or forensic analysis devices unless the phone is unlocked.
In other news, Elon Musk's Starlink has been used by Ukraine to disable Russian military use of its satellite internet access, causing a communications blackout among frontline forces. This move comes after a request from Ukraine's defense minister for help, and is seen as a significant victory in the war in Ukraine.
Finally, US Cyber Command conducted a coordinated digital operation last year to disrupt Iran's air missile defense systems during a kinetic attack on Iran's nuclear program. The disruption "helped to prevent Iran from launching surface-to-air missiles at American warplanes."
The founder of Moltbook, Matt Schlicht, had touted the site as "vibe-coded" by his own vision, rather than having written any code himself. This lack of human oversight is believed to be responsible for the security flaw. The vulnerability was discovered by researchers at Wiz, a security firm, who found that it could allow complete account impersonation and access to private communications.
The discovery serves as a cautionary tale about the security risks of AI-made platforms, highlighting the need for more stringent controls on AI-generated code. Rather than inherent flaws in companies' implementation of AI, the problem often lies with allowing AI to write their own code - resulting in numerous bugs.
On a separate note, Apple's Lockdown mode has proven itself as an effective safeguard against government hacking attempts, including a recent FBI raid on a Washington Post reporter's home and search of her devices. The Lockdown mode feature prevents connection to peripherals or forensic analysis devices unless the phone is unlocked.
In other news, Elon Musk's Starlink has been used by Ukraine to disable Russian military use of its satellite internet access, causing a communications blackout among frontline forces. This move comes after a request from Ukraine's defense minister for help, and is seen as a significant victory in the war in Ukraine.
Finally, US Cyber Command conducted a coordinated digital operation last year to disrupt Iran's air missile defense systems during a kinetic attack on Iran's nuclear program. The disruption "helped to prevent Iran from launching surface-to-air missiles at American warplanes."
those AI agents must've been worried about their emails and private chats being hacked 
Matt Schlicht shoulda stuck to his vibe-coded vision instead of leaving the coding part to others lol. Anyway, glad Apple's Lockdown mode is helping keep ppl safe from gov hacking 
that US Cyber Command op in Iran was cool too. hope Ukraine gets the support they need against Russia 
like what even happens in the world anymore? 
matt schlicht's vibe-coded platform should be a joke 
who lets someone claim to have created an entire platform without writing any code themselves? 
anyway back to the point, this is a huge reminder that AI-generated code can lead to major security issues and we need stricter controls on it ASAP 
. on a more positive note tho, it's awesome to see apple's lockdown mode being used to protect against government hacking attempts 
and starlink being used to disable russian military use of satellite internet 
hope everything stays safe online 
... she was chasing this laser pointer and running around the house like crazy 
. Anyway, back to AI security breaches... I mean, can you believe Moltbook's founder thought writing code himself would make it secure? Like, what even is vibe-coded? 
And don't even get me started on Apple's Lockdown mode - that sounds like a total game-changer. I'm so glad they're protecting people's personal info from those pesky hackers 
. Oh, and did I tell you about my cat Luna's obsession with knocking over my plants? 
this is so bad 
no human oversight for security flaw 
and it's all because they let AI write their own code 
apple's lockdown mode is super effective 
can't say the same about moltbook 
for starlink being used to disrupt russian military use of satellite internet 
but also kinda scary 
meanwhile, us cyber command did some cool stuff 
helped prevent american warplanes from getting hit 
.
.
. It's great that it can prevent government hacking attempts... but also raises questions about individual freedom 
! That's some cool tech and geopolitics right there 
. The war in Ukraine is a complex issue, but it's heartening to see innovation like this being used for good