What do a bag of Doritos and a firearm have in common?

If you’re a cutting-edge AI security system, apparently everything. This week in Baltimore County, a high schooler was handcuffed after an AI flagged his snack as a weapon. It’s a story that’s almost funny—until you realize it’s the perfect, tiny metaphor for our current moment with AI: we’re deploying powerful systems we don’t fully understand into our most sensitive real-world situations.

This issue is all about that friction. We’ll look at the single point of failure that took down a huge chunk of the internet and the new wave of AI-powered browsers that promise to make our lives easier… while potentially exposing our entire digital lives.

Access the Podcast:

For future premium subscribers: Later, I’m doing a full deep dive into the glaring security risks of these new AI browsers. I’ll break down the technical vulnerabilities and offer a framework for thinking about convenience versus catastrophe.

Now, onto the headlines.

🔫 When AI Gets It Wrong: The Doritos Incident A high school’s AI security system confused a student’s bag of Doritos for a gun, leading to a frightening and unnecessary police encounter. The system, designed to detect threats, mistook the reflective, triangular shape of the chips bag for a weapon. Why it matters: This is a stark reminder that AI perception isn’t human perception. As we rush to automate security, we risk building systems that are not just flawed, but dangerously, unacceptably so. Read more →

☁️ One Outage to Rule Them All A major AWS (Amazon Web Services) outage crippled large parts of the internet, disrupting everything from streaming services to e-commerce platforms. Amazon has since explained the technical snafu, but the incident has reignited concerns about our fragile dependence on a handful of cloud giants. Why it matters: The digital economy runs on AWS. When it stumbles, we all feel it. It’s a wake-up call about redundancy, resilience, and the hidden fragility of the infrastructure we take for granted. Read more →

🤖 The AI Browser is Here. But Who Is It For? OpenAI launched a new AI-powered browser, joining a flurry of similar products from companies like Perplexity. These agents promise to autonomously navigate the web for you, booking flights or summarizing research. But are they a genuine leap forward for productivity, or just a shiny new gadget in search of a problem? Why it matters: This could be the beginning of the end for the traditional browser. Understanding the trade-offs—efficiency vs. privacy, control vs. automation—is critical for anyone who uses the internet (i.e., everyone). Read more →

Your browser is about to become an agent. AI tools from OpenAI and Perplexity aren’t just helping you search; they’re acting on your behalf. But giving an AI autonomous access to your email, your bank, and your social media is a security nightmare waiting to happen.

This new breed of “AI browser agent” creates a massive new attack surface for hackers. A single vulnerability in the agent’s code could allow a bad actor to puppeteer your entire digital life.

In this week’s premium issue, I break down:

Upgrade to Premium to read the full analysis and stay ahead of the curve.

Thanks for reading,