TrendNew Politics. Diplomacy. Markets. Tech. What matters.
Tech 6 min read

The Summer of Breaches: When America's Tech Stack Started Cracking

From hacked routers to quantum threats to AI music spam—the infrastructure we all depend on is breaking in real time. And nobody's really ready.

The Summer of Breaches: When America's Tech Stack Started Cracking

The FAA just grounded Blue Origin’s New Glenn rocket. Iran-linked hackers just hit US critical infrastructure. Russian military just pwned thousands of consumer routers. And somehow that’s not even the worst news this week.

Welcome to the summer where America’s tech stack stopped pretending to be secure.

When the Attacks Get Boring Because They’re Constant

Let’s start with what should terrify you but probably won’t make the evening news: thousands of consumer routers across the US got compromised by Russia’s military. Not some script-kiddie operation. Not ransomware gangs looking for quick ransom. The actual Russian military decided your neighbor’s Wi-Fi router was worth hacking.

This isn’t sophisticated. It’s not even particularly novel. It’s just scale. The same vulnerability patterns we’ve known about for years, exploited at mass. It’s like if someone discovered your front door lock was broken and instead of fixing it, the Pentagon just started walking through everyone’s houses to see what was on the coffee table.

Then you’ve got Iranian hackers disrupting operations at US critical infrastructure sites. Again—not the dramatic, movie-plot kind of disruption. Real disruptions. Real infrastructure. Real people whose power or water or networks got dinged because adversaries decided Tuesday was the day.

Meanwhile, a currency exchange sanctioned by the US just lost $15 million to “unfriendly states.” That’s the official language they used. Not “unknown threat actors.” Not “sophisticated hackers.” Just: yeah, probably China or Russia, we’re not entirely sure, fifteen million gone.

Scenic sunset at Ipanema Beach, Rio de Janeiro, with crowds and vibrant atmosphere. Photo by Iryna Olar / Pexels

The Quantum Elephant That’s Already In The Room

Here’s where I get genuinely concerned: Big Tech is now officially in the “Q-Day danger zone.”

Recent advances in quantum computing have pushed us past the theoretical stage and into the “this might actually break everything” stage. And I don’t mean “break” in the sense of “needs a software patch.” I mean break like “every encrypted file your bank has ever stored could be retroactively decrypted if someone saves the encrypted data now and waits five years for quantum computers to mature.”

The NSA knows this. They’re so worried they’re actually using Anthropic’s restricted Mythos AI model—despite the fact that there’s an open feud between the Pentagon and Anthropic. When bureaucracies overcome institutional beef to collaborate on something, you know they’re scared.

My read: we’re watching the intelligence community silently admit that quantum threat timelines just compressed. And they’re not going public about it yet because the market doesn’t need that kind of panic.

The Enterprise Revolt Nobody’s Talking About

Here’s something weird that actually tells you something about the state of Big Tech right now: thousands of companies are in the middle of migrating away from VMware because of “negative” views of Broadcom, which acquired VMware.

This is quiet rebellion. Not a dramatic hack, not a security breach—just an unglamorous, expensive infrastructure overhaul driven by corporate skepticism. When enterprises start migration projects this size, it’s because trust is gone. You don’t rip out your entire virtualization layer for fun.

What does that tell you? It tells you that the relationship between Big Tech vendors and their enterprise customers has fundamentally corroded. It’s not about security theater anymore. It’s about “we don’t trust you to not squeeze us next quarter.”

Close-up of hands holding a smartphone displaying 'Announcing Grok 3' on a dark background. Photo by UMA media / Pexels

The AI-Generated Music Scam Is Already Here

Deezer just reported that 44% of songs uploaded to its platform daily are now AI-generated. That’s not 44% of total library. That’s 44% of new daily uploads.

Here’s the trick though: consumption of AI-generated music is only 1-3% of total streams. Why? Because 85% of the AI-generated music streams are fraudulent and demonetized. So there’s this weird equilibrium where the platform is being flooded with bot-generated spam that doesn’t actually make money but does clog the system.

It’s like if your email inbox was suddenly 44% spam, except the spam doesn’t even convert to clicks. It’s just noise designed to game metrics and confuse the algorithm.

This is what happens when you make it trivially easy to generate content and you have financial incentives to game the system. The incentives don’t actually need to be good—they just need to exist. Some actor somewhere figured out that machine-generated music plus bot streams equals maybe a penny or two, multiplied by volume.

It’s not sophisticated. It’s just what happens when the barrier to entry disappears.

What Actually Worries Me

The thing that keeps me up isn’t any single incident. It’s the pattern.

We’ve got state actors running garden-variety attacks at scale (routers, infrastructure). We’ve got quantum threats compressing into timelines we can’t publicly discuss. We’ve got enterprises quietly abandoning vendors they don’t trust. We’ve got AI spam flooding music platforms. We’ve got a rocket grounded. We’ve got currency exchanges losing millions.

None of these are 10x problems individually. But together? This looks like the year the bill came due for two decades of building fast, shipping broken, and fixing it in production.

My prediction: we’re about to see a hard pivot toward “boring security” across enterprise. Not flashy zero-trust frameworks or whatever gets pitched at conferences. I’m talking about the expensive, painful work of actually auditing what you’re running, where your data lives, and whether you can trust your vendors. The unglamorous stuff that should’ve happened in 2015.

The companies that started this work three years ago will be fine. The ones waiting until their board gets breached will be in for a bad time.

Here’s what I genuinely don’t know: whether this is a inflection point or just a normal summer of chaos in a permanently chaotic industry. I’ve covered this beat for 12 years and I honestly can’t tell if we’re watching a systemic reset or just pattern noise.

But I think the smart money is on “reset.”

What I’m Watching

  • Blue Origin’s New Glenn investigation outcome, specifically whether the FAA’s findings reveal design problems or one-off manufacturing issues. If it’s design, this delays commercial operations by years. If it’s manufacturing, we’re back on schedule in weeks. The difference is whether SpaceX keeps the monopoly on heavy lift or actually has competition by 2026.

  • Enterprise migration velocity from VMware/Broadcom. Track quarterly filings from mid-market cloud providers and watch if they report new customer wins explicitly mentioning “alternative to Broadcom stack” language. If migrations accelerate beyond current projections in Q4, you’re watching a vendor relationship actually die.

  • NSA quantum roadmap leaks. Someone’s going to leak the actual timeline on when they think quantum decryption becomes real. Watch for “document obtained by” reports in the next 6-12 months. When it drops, that’s your signal for how much time we actually have to re-encrypt everything.

  • Whether Deezer or similar platforms actually implement friction against AI music uploads. If they don’t within Q4, you’ll know they’ve decided bot-flooding is cheaper than moderation, and it’ll spread to every platform that monetizes on volume.