Behind the video
The AI Hacks Have Begun. What Happens Next?
Overall Thesis
The video is a documentary-style explainer on how AI has accelerated both offensive cyberattacks and defensive cybersecurity tooling through October 2026, citing breaches at Binance, Chainalysis, and others, and sketching four possible futures for the industry. It does not advance an investment thesis or price prediction for any named company.
Narratives
CRWDCrowdStrikeCrowdStrike is discussed as one of several major security vendors selling AI agents that triage alerts and investigate threats, and is referenced in connection with the 2024 outage caused by a faulty update. The host does not express a view on CrowdStrike's stock price or investment merit.
Key Arguments
- CrowdStrike tracked a Chinese state-backed group running the Artex AI tool against more than 200 organizations.
- CrowdStrike, along with Microsoft, Google, and Palo Alto Networks, now sells AI agents that sort alerts, investigate, and act autonomously.
- A CrowdStrike spokesperson is quoted discussing the need for fast, tactical autonomous defense without a human in the loop for every decision.
- The July 2024 CrowdStrike update crash that grounded flights and canceled surgeries is cited as a cautionary example of automated systems acting at scale without human review.
Risks acknowledged
- In an industry survey, not a single team let its AI defender act fully autonomously, with most requiring human review of every decision.
GOOGLAlphabetGoogle is discussed for its Mandiant threat intelligence data, its Claude-Mythos-style vulnerability discovery efforts, its $32 billion acquisition of cloud security firm Wiz, and its AI-based spam filtering in Gmail and Chrome. No investment thesis or price view is expressed.
Key Arguments
- Google's Mandiant reported that the average time for attackers to exploit a disclosed vulnerability has fallen to -7 days (attackers often exploit before disclosure).
- Google paid $32 billion for cloud security company Wiz, described as the biggest deal in Google's history, consolidating cloud security into a few large providers.
- Google already has an AI sending fixes into open-source projects, described as potentially 'the most important idea in this whole story'.
- Gmail blocks more than 99.9% of spam using machine learning, offered as a model for how AI-driven cybersecurity could eventually make hacking 'constant, automated, and mostly invisible'.
Risks acknowledged
- Consolidating security into a handful of cloud providers means that when one of them has a bad day, everyone has a bad day.
Hedges & Caveats
- We looked for a documented case of a defensive AI stopping a live AI attack. We couldn't find one.
- It's early, but it may be the most important idea in this whole story (regarding AI auto-patching).
- Illustration animations are simplified explanations, not real logs or screenshots.
- All content is for discussion and illustrative purposes only and should not be construed as financial advice.