Tech's Trust Collapse Is Getting Harder to Hide

---

Sam Altman apologized. Not for anything OpenAI did wrong operationally—for not telling police about a mass shooting suspect’s account fast enough.

The letter to Tumbler Ridge, Canada sits in that weird space where corporate contrition meets admission of failure. Altman wrote he was “deeply sorry,” which is the bare minimum acknowledgment when your platform was involved in something that got people killed. What’s wild isn’t that he apologized. It’s that this feels like the tenth domino in a very specific pattern nobody’s quite naming yet: tech leadership has stopped even pretending to have mechanisms for accountability.

Start here. OpenAI didn’t just fail to report the suspect’s activity to authorities—they had to be asked by journalists or regulators before acknowledging it happened at all. That’s not a speed problem. That’s a culture problem.

Photo by Rostyslav / Pexels

The Accountability Vacuum

Meanwhile, Sam Altman’s simultaneously trying to make OpenAI actually profitable for the first time. The company’s culling projects, getting “more disciplined with strategy”—which is corporate speak for “we spent money on stuff that didn’t return value.” That’s fine. Companies do that. Except Altman’s also the guy who couldn’t be bothered to notify police about a mass shooting suspect, which suggests his discipline is selective.

The timing matters. You can’t be the visionary CEO pushing AI safety while simultaneously showing that your own company’s operational discipline doesn’t extend to “inform law enforcement about shooters.” These aren’t parallel concerns—they’re the same concern. If you can’t build systems that catch a mass shooting suspect using your platform, what exactly are you building systems for?

Then there’s Elon.

SpaceX, by all appearances, has become a financial instrument for Elon Musk to use. A New York Times examination found the rocket company gave Musk loans, propped up his other businesses, and functioned less as a standalone operation and more as a subsidiary of his personal balance sheet. This isn’t illegal. It’s not even unusual among founder-led companies. But it’s the opposite of what we’ve been sold. SpaceX was supposed to be the validation of the free market, the private company that could do what NASA couldn’t. Instead it’s a piggy bank.

The pattern across these stories is identical: structure enables opacity, and opacity enables things that wouldn’t survive sunlight.

Photo by nappy / Pexels

When Governance Becomes Theater

A tech CEO just published a 22-point manifesto about the future of the West. He runs a company with NHS and defense contracts. The CEO doesn’t need to convince shareholders or employees—he just drops ideology directly onto the internet and moves on. That’s the dream outcome for someone who’s made it clear they don’t want constraints.

Meanwhile, the UK’s National Cyber Security Centre is pushing for passwords to die, replaced by passkeys. Good idea, genuinely. But who’s enforcing that transition? Who’s making sure the systems that replace passwords don’t become the next attack surface? Nobody’s asking because nobody’s accountable yet.

A biobank had a data incident. A professor running it said “a few bad apples” were responsible. That’s real accountability theater right there. A few bad apples imply isolated problems. Isolated problems don’t require structural change. Isolated problems are tragic but manageable. Except biobank data is biological data—it’s not like a database breach where you reset passwords. You can’t un-know someone’s genetic information. The incident happened. The framing of it as rogue employees rather than systemic failure is gaslighting with a British accent.

Here’s my honest uncertainty: I don’t know if this is a turning point or if it’s just what the system looks like when it’s scaling. Silicon Valley’s always had credibility problems. But there’s something different about the current moment. The failures aren’t hidden in the fine print anymore. They’re public. And the responses aren’t defensive—they’re barely responses at all.

The Information Chaos Problem

When gunfire erupted at the White House Correspondents’ dinner on Saturday, journalists and guests posted what they experienced in real time. Confusing, contradictory, human accounts. Within hours, influencers filled the gaps with conspiracy theories. Nobody was accountable for the theories. Nobody had to be. The gap between what happened and what people believed had expanded faster than any official communication could close.

This is the actual innovation that tech platforms enabled: the ability to scale speculation faster than facts can move. OpenAI’s silence about the Tumbler Ridge shooter? That silence didn’t last—it just got filled with whatever people wanted to believe about it. SpaceX’s financial relationship with Musk? That wasn’t hidden for long—it just took a Times investigation. The biobank incident? Now it exists in that zone where some people believe it was bad security, others believe it was intentional, and nobody from leadership is moving fast enough to make that zone smaller.

My prediction: This gets worse before it gets better, and I don’t think anyone’s built the infrastructure to handle it.

Tech companies have regulatory compliance. They’ve got ethics boards and safety teams and incident response protocols. What they don’t have—what they’re actively avoiding—is accountability that costs something. Altman’s apology cost him nothing. The manifesto CEO has government contracts and apparently no mechanism to revoke them based on ideology. SpaceX keeps functioning. The biobank keeps operating.

There’s a reason nobody’s fixing this. Because the moment you build actual accountability, you have to accept that it applies to you too.

What I’m Watching

• OpenAI’s next incident timeline. If there’s another content policy failure (terrorist content, CSAM, whatever) between now and March 2025, watch whether Altman’s “discipline” includes transparency improvements or just PR. Specific trigger: does OpenAI publish a responsible disclosure policy within 60 days?

• SpaceX loan patterns. Track whether Musk uses SpaceX as collateral for other ventures after the Times story. If he does it again visibly, that’s the moment the regulatory temperature has to rise. The SEC or Congress will have to explain why a government contractor can function as a personal asset vehicle.

• The passkey adoption cliff. The NCSC said passwords should die. By Q2 2025, watch which major platforms actually implement passkey-only options. Whoever doesn’t is either technically stuck or strategically delaying (which means they’re not confident in the security). That gap tells you who’s actually serious about the security shift versus who’s just following headlines.

• Biobank legislation. The incident was real. Watch whether the UK pushes for legal liability for data custodians rather than the “bad apples” framing. If legislation happens by summer 2025, institutions took it seriously. If not, they got away with reframing.

The real watch: can accountability scale as fast as the failures?