2015-era Google Security swag from my collection

Adult Europeans with a social media permit are going to be so mad at you

No, your tax dollars not paying for this. The Ad Council is a nonprofit that promote public-service announcements. The initial stated reason was to get nice charitable messages out to the public, but now it's pretty much just used to fill unsold ad time. These are wildly common on overnight radio. Now some of the ads might have been paid for by the government - the NHTSA runs *a lot* of traffic-safety ads - but they're not paying for any airtime.

Taylor Swift observing her children at the Game Day Rec dance regionals

Scored these at work today

Can someone in American manufacturing explain if this factory is... real? How is this amount of labor and speed profitable? youtu.be/p40gaCou2Qs

IT HAPPENED. I pointed at a thing and called it a tank and my partner was like, "um, no, it's a self-propelled howitzer"

Saw someone driving my car except with their top on and I let them know exactly how much of a piece of shit I thought they were

RIP to all the baddies lost on Olympic Carrier, this date in 21354

RIP to all the baddies lost on the Olympic Carrier, this date in 21354

Reminds me of my early PFE days when I had to fly to Alaska last minute because a customer was experiencing AD authentication issues. I get there, nothing wrong with the DCs, but nothing outside the datacenter can auth. Run packet captures, smells like network issue. Customer swears it isn't. On a hunch, I ask them if they recently rotated keys on a TACLANE (classified network), they say no, but they just upgraded one right before the issue started. Lucky for them I was former USAF and still in the ANG at the time, with experience managing TACLANEs. I tell them to rotate keys on both ends. Problem solved. Customer POC is super embarrassed for calling in a crit-sit when it wasn't remotely a MSFT problem.

NEW: malware developers added nuclear & biological weapons text to to their spyware. Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner. Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky. When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit. We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted. In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation. H/T to colleagues that shared this with me socket.dev/blog/mini-shai-hu…