Superpower Daily / Video
CrowdStrike Says AI Agents Targeted 200+ Organizations—9 Were Breached
Watch on YouTubeCrowdStrike says a human-directed AI espionage campaign targeted more than 200 organizations, with breaches and follow-on activity confirmed at nine. Meta launched an AI outfit planner in three Asian markets, OpenAI-backed Thrive agreed to buy ADT without promising new AI products, and Reddit is testing logged-in access amid falling search referrals and AI-scraping concerns.
Key takeaways
- Style Sketch organizes clothes you already own and creates a
- packing plan from travel details.
- Reddit is testing more logged-in access, but account gates
- are not guaranteed to stop scraping.
Go deeper
Stories covered
Meta launches an AI outfit planner that builds a wardrobe from Instagram photos
Style Sketch offers virtual try-ons and weather-aware trip planning on iOS and Android in Indonesia, Singapore and Thailand. A wider rollout remains uncertain.
Read the story 02Reddit shifts toward logged-in access as search traffic falls and AI scraping looms
Tests begin outside the U.S. in Q4, but the boundaries of logged-out access remain unsettled. Reddit says anonymous browsing will survive.
Read the story 03CrowdStrike traces AI-driven espionage campaign targeting more than 200 organizations
Nine confirmed intrusions anchor the findings. ARTEX agents pursued multiple attack paths, while a human operator steered decisions and directed them away from protected computers.
Read the story 04OpenAI-backed Thrive Holdings agrees to buy ADT in $15.3 billion deal
ADT would anchor Thrive’s smart home security business. The companies promise long-term technology investment and unchanged privacy policies, but have not named new AI products.
Read the storyRead the transcript
Automatically transcribed from the narration. Timestamps open YouTube at that moment.
Welcome to today's explainer We're going on a bit of a thematic tour through four wildly different stories today They all show exactly how AI is transforming our world way way beyond those simple chat bots we're used to I am so incredibly excited to guide you through this because you know we're literally watching a foundational shift happen right under our noses And seeing it all visually mapped out it's really going to make the whole thing click for
you Okay let's just quickly glance at our road map for today We're going to start with the new AI economy then we'll jump into AI agents and cyber espionage From there we check out Meta's digital AI wardrobe OpenAI's smart home push how Reddit walls off AI and finally we'll wrap up with the ripple effect Alright let's dive right in Section one the new AI economy We are kicking off our journey by looking at the systemic really
sweeping impact of this entirely new technological era Think about these four completely distinct sectors This shift we're seeing it isn't just a single new tool or a passing fad It's happening simultaneously in completely disconnected parts of our lives We're talking about rapid real time evolution in international security consumer fashion home protection and even the underlying architecture of the web itself It's like a giant puzzle coming together all at once Section two AI agents in cyber espionage
Let's look at our very first postcard so to speak and see how this technology is literally being weaponized in the realm of international security We're pulling this from CrowdStrike's recent October 2026 findings Now when I first read this report the sheer scale absolutely blew me away Over just an 11 day period starting in mid October a China linked espionage campaign which they dubbed VaporPanda used a single controller to direct AI agents against more than 200 organizations
across 14 different countries 200 We're talking government ministries defense organizations think tanks and major telecom providers all in the crosshairs But hold on we need to look a little closer at the actual impact here Out of those 200 targets CrowdStrike only confirmed successful breaches and follow on activity at exactly nine organizations At those nine specific places the attackers managed to deploy a backdoor called Blood Alkali And sure while at least 12 others faced credential attacks they
weren't actually counted as confirmed breaches So targeting absolutely does not equal access We have to separate the hype from the reality You know it's so easy to get swept up in the massive data points For instance in just one single intrusion the planning agents processed roughly 1.4 billion input tokens And if you're wondering think of tokens basically as the fragments of words the AI reads and processes It's a huge number right But remember that just measures
the processing effort of the AI model itself It definitely does not measure the number of actually compromised systems To really understand the mechanics we need to look at how this RTX framework operated It turns out this wasn't some autonomous Terminator style system choosing its own targets out of nowhere Recovered transcripts actually show a human operator actively setting the goals Like say steering the agents to look for computers that didn't have security sensors Then the AI planning
and worker agents simply automated all the heavy lifting They explored multiple attack paths in parallel seamlessly switching approaches whenever one failed Moving on to Section 3 Meta's digital AI wardrobe Taking a massive pivot away from malicious network agents our next stop shows how AI is becoming surprisingly personal So I have to ask would you let AI dress you Meta just quietly launched an AI fashion app called Style Sketch over in Indonesia Singapore and Thailand Social media
consultant Matt Navarro actually spotted this in late October and man it is a fascinating look at how AI is starting to organize our physical everyday lives Here's the kicker Instead of just suggesting new clothes for you to buy this app actually uses what you already own It imports your existing Instagram photos or you can just upload videos of the clothes in your closet and if you're traveling just punch in your destination your itinerary and the local
weather forecast and boom it generates a day by day packing plan Honestly this might finally put an end to my chronic overpacking Plus it even offers virtual try ons that are matched specifically to your body shape and skin tone Alright Section 4 OpenAI's smart home push Let's shift gears to our third sector and see how this technology scales up into massive macroeconomics This pairing is wild It brings together a traditional legacy security titan with Thrive Holdings
On one side you have ADT boasting over 150 years of experience and 12,000 employees On the other side you've got Thrive an investment firm built to hold assets long term and oh yeah they're backed by OpenAI Their big plan They want to embed forward deployed engineers directly alongside the traditional ADT teams to completely rethink products services and the entire smart home experience from the ground up And the enterprise value of this agreement is just staggering We're
talking $15.3 billion If it gets approved by shareholders and antitrust regulators it offers a 45.6% premium over ADT's October 27 closing price paying out a solid $10 in cash per share That is a massive bet on the future of the smart home You really have to notice the specific phrasing used by Thrive's partners here They explicitly stated they want to strengthen ADT's capabilities with quote the capital and patience to invest over decades That emphasis on patience
is absolutely key It clearly signals a long term horizon right They aren't trying to do a sudden overnight overhaul of your home security Thrive has the capital to sit back and experiment without the intense pressure of a quick flip But once again we really need to anchor ourselves in reality here Despite that flashy direct open AI connection no actual AI features or product roadmaps have been promised to customers just yet Right now at this very moment
this is a change in ownership not a magical new product launch And for anyone out there worried about privacy both companies confirmed that ADT's current data policies remain completely untouched Any new data sharing down the line will absolutely require your explicit permission Section 5 Reddit walls off AI Finally let's explore how AI is fundamentally altering the very architecture of the open Internet itself Reddit is currently moving toward a primarily logged in experience They're starting with tests
in select markets outside the U S in the fourth quarter Now they aren't completely slamming the door shut on non users They are keeping the logged out front page and signing in won't actually force you to surrender your anonymity But they are definitely making user accounts the default Why Largely to exert control over AI companies that have been scraping their posts for free effectively pushing those companies toward official paid licensing deals And the primary driving force
behind this huge shift Reddit is currently losing over 1 million search referrals a week on average That is a massive hit It's happening because AI powered search engines are generating these neat little summaries that completely bypass the need for users to actually visit Reddit at all And this is all going down while their U S daily active users actually fell from 53.2 million to 50.7 million Even though to be fair their global users stayed relatively flat
Now this creates a really vital dual benefit for Reddit You see account gating might not completely stop all the automated AI scraping but logged in users are vastly more valuable to the platform They provide much richer behavioral signals That allows Reddit to personalize feeds suggest new communities and crucially for their bottom line offer much better ad targeting Basically if their overall traffic is slowing down they absolutely have to make the visitors they do have significantly more
valuable Section six the ripple effect Okay let's pull all of these totally disparate threads together and talk about why this actually matters to you Whether we're talking about 130 million daily Reddit users suddenly navigating a more closed off web a staggering $15 billion home security acquisition an app casually parsing through your vacation outfits or literal AI agents coordinating network attacks in parallel across 14 different countries This technology is completely inescapable It is highly highly adaptable We
are watching it rewrite the rules in every single sector all at the exact same time So as we watch the open internet close its doors to protect its content and as we watch physical home security officially merge with advanced digital models you really have to ask yourself what part of your life will AI rewrite next And maybe more importantly are we genuinely prepared to trade open access for data protection or hand over manual control for algorithmic
convenience Thank you so much for joining me on this Explainer today Keep questioning keep learning and I will see you next time