The World Is F@cked #3: National Security Until the AI Lawyers Arrive

AI became a military asset, a consumer gadget, a courtroom exhibit, and a quarterly sales target at the same time.

Image from Pexels

Everyone is technically correct, but they’re pulling in different directions, creating confusions

I have reached the point where I no longer know whether to read the technology pages, the business pages, the legal pages, or the comedy section. They all seem to be covering the same types of stories. I thought Apple and OpenAI were friends, but apparently they’re now rivals in court.

One day, artificial intelligence is declared a matter of national security as I covered in the Antropic story before. The next day it is marketed as a friendly assistant that writes shopping lists and holiday itineraries. Then a courtroom fills with lawyers arguing over trade secrets, while investors ask the only question that truly matters on Wall Street: “How much revenue did it generate this quarter?”

Governments say AI is a national security asset. Companies say AI is a commercial product. Lawyers say AI is intellectual property. Investors say AI is quarterly revenue. Engineers say AI should be open. Military planners say AI should be controlled. Everyone is technically correct. And everyone is pulling in a different direction. That, to me, is the real story.

Somewhere between geopolitics, billion-dollar lawsuits, export controls and polished aluminum, I found myself thinking only one thing: The world is officially f@cked.

I feel this way not because artificial intelligence has become too intelligent or humanity has become magnificently contradictory. But consider the headlines that have appeared within just a few weeks after the Anthropic saga I shared with you in a few posts earlier.

Governments warn that frontier AI models are so strategically important they require export controls. Companies assure us AI will make humanity healthier, wealthier and more productive. Apple sues OpenAI, alleging confidential hardware knowledge followed former employees into one of Silicon Valley’s most ambitious AI hardware projects.

Those allegations remain to be tested in court, but the lawsuit illustrates how valuable engineering know-how has become. Meanwhile, reports indicate that subsidiaries of Chinese technology companies operating outside mainland China have legally accessed advanced AI services in accordance with existing regulations.

You can read that last sentence again. It is not because anyone necessarily broke the law, and they apparently didn’t. That is the delicious irony.

The same technology can be described as a strategic military capability, a cloud subscription, a consumer product, intellectual property, and a quarterly revenue stream, all before lunchtime. Apparently, the classification depends less on the technology than on which meeting you’re attending.

If you are in Washington, AI is national security. If you are in Silicon Valley, AI is innovation. If you are in a courtroom, AI is evidence. If you are on an earnings call, AI is next quarter’s guidance. And if you are an ordinary citizen trying to understand what is going on, AI resembles a plot twist written by someone who misplaced the script halfway through the film.

Then came the developers’ concerns about Grok today! It was just when I thought the plot couldn’t get any stranger that another headline wandered onto the stage.

A security researcher reported that xAI’s Grok Build CLI appeared to upload entire Git repositories — including, in some cases, unredacted .env files that may contain API keys and other sensitive credentials—even when only a small portion of the code was needed.

The researcher also claimed that turning off the data-collection option did not initially prevent those uploads. xAI subsequently changed the service behavior through a server-side update, and Elon Musk later said the collected data would be deleted.

Whether that incident ultimately proves to be a design flaw, an engineering oversight, or simply a lesson learned, it highlights a broader issue that extends far beyond one company.

Artificial intelligence asks us to entrust it with the most valuable assets we possess, such as our code, unpublished research, intellectual property, and business plans. Sometimes, literally, the keys to our digital kingdom.

As someone who spent more than five decades in science and public health, I was taught a principle that remains timeless: extraordinary claims require extraordinary evidence. Perhaps the AI era needs a companion principle. Extraordinary convenience requires extraordinary verification.

The old computing adage, “Never run software you don’t trust,” has evolved into something far more challenging: Never assume an AI assistant only sees what you intended it to see.

That isn’t a criticism of one company. It is a reminder that trust should never be blind, especially when software can move information at the speed of light. And this is where our modern AI world becomes almost comically contradictory.

Apple is reportedly worried about hardware secrets. Governments are worried about military secrets. Developers are worried about source-code secrets. Lawyers are worried about trade secrets. Investors are worried about market share. Meanwhile, the rest of us are simply trying to remember which chatbot we’ve just given access to our documents.

If this were a Hollywood screenplay, critics would probably reject it for being too implausible. Reality, however, does not worry about believable plots.

Perhaps that is the defining characteristic of our age. The technology is revolutionary. The incentives are contradictory. The laws are struggling to keep pace. And the rest of us are left wondering whether the future is being designed by engineers, regulators, shareholders or lawyers. The answer, of course, appears to be yes to all.

Well while dealing with these AI issues I also swallowed my my expensive pill and turned it to juicy yar. Here is the link if you wish to check out.

[embed]The Most Expensive and Tiniest Pill I Have Ever Swallowed in the Last 79 Years A Single $66 Heart Pill Originated from France: How One Tiny Tablet, Coralan, Exposed the Hidden Economics of Modern…medium.com

Cheers,

Mike