Forgot your password?
typodupeerror

Submission + - How Microsoft's "Little Workaround" Created a Major Pentagon Threat (propublica.org)

joshuark writes: ProPublica Reporter Renee Dudley heard Microsoft was running tech support for the U.S. Defense Department through China, the country’s biggest cybersecurity adversary.

The arrangement was called “digital escorting.” She thought it sounded like a conspiracy theory — until she started looking into it. This is the story of what she found and how her investigation changed government policy.

Microsoft is using engineers in China to help maintain the Defense Department’s computer systems — with minimal supervision by U.S. personnel — leaving some of the nation’s most sensitive data vulnerable to hacking from its leading cyber adversary, a ProPublica investigation has found.

The arrangement, which was critical to Microsoft winning the federal government’s cloud computing business a decade ago, relies on U.S. citizens with security clearances to oversee the work and serve as a barrier against espionage and sabotage.

National security and cybersecurity experts in the Trump administration contacted by ProPublica were also surprised to learn that such an arrangement was in place, especially at a time when the U.S. intelligence community and leading members of Congress and the Trump administration view China’s digital prowess as a top threat to the country.

Microsoft uses the escort system to handle the government’s most sensitive information that falls below “classified.” According to the government, this “high impact level” category includes “data that involves the protection of life and financial ruin.” The “loss of confidentiality, integrity, or availability” of this information “could be expected to have a severe or catastrophic adverse effect” on operations, assets and individuals, the government has said. In the Defense Department, the data is categorized as “Impact Level” 4 and 5 and includes materials that directly support military operations.

“If someone ran a script called ‘fix_servers.sh’ but it actually did something malicious then [escorts] would have no idea,” a former Microsoft engineer who worked on the escort system, told ProPublica in an email. That said, he maintained that the “scope of systems they could disrupt” is limited.

In an emailed statement, the Defense Information Systems Agency said that cloud service providers “are required to establish and maintain controls for vetting and using qualified specialists,” but the agency did not respond to ProPublica’s questions regarding the digital escorts’ qualifications.

It’s unclear whether other cloud providers to the federal government use digital escorts as part of their tech support. Amazon Web Services and Google Cloud declined to comment on the record for this article. Oracle did not respond to requests for comment.

A spokesperson for the inspector general — whose office is supposed to operate independently in order to investigate potential waste, fraud and abuse — told ProPublica they were not authorized to speak about the issue and directed questions to DISA public affairs.

Comment Fundamental mismatch? (Score 2) 44

But why should you focus on the US? But my main reaction to that specific data is that Americans seem to have quite limited choice in the smartphone market.

I do thank the OP/FP poster for saving me the websearch. However I feel like I should have had some awareness of the OnePlus brand if only because I'm using my second Oppo smartphone now...

From that perspective it sounds to me like a fundamental mismatch. I see Oppo as a commodity brand. Low cost, low profit, satisfactory value. But OnePlus was apparently going for the high end market. But my own mistake because I'm generally shopping at the low end and expecting to soon discard the phone for better tech. I sometimes ask the salespeople to explain why I would consider paying five or 10 times more for a fancy smartphone and so far none of them have managed to pique my interest.

(And I still think Huawei was the best value from the consumer perspective--but without knowing how much of my personal data got compromised. (No problem until I start thinking (or exposing) bad thoughts about Xi?))

Comment Imploding or exploding? Which is funnier? (Score 1) 46

"Missed it by so close!"

– Obligatory Mariner Smart quote

Don't you mean "Missed it by that much"? And Max Smart?

However the reference is still weak because he was using his hand to indicate a small distance, which should be a "this" reference. Where is "that" supposed to be pointing?

I think we had better jokes in those days. Possibly somehow related to IBM's days of glory? I still meet or hear from some Big Blue folks from time to time. I don't think any of us can understand what happened to the company...

As regards this story, it will only be funny if IBM somehow helps cause the AI bubble to burst or implode. I'm still not sure which way it will go. To burst would imply there is something inside there, perhaps a dangerous AI that trying to capture all the world's wealth overnight? Implosion would imply there's nothing there and the "financial container" finally got too weak relative to the "value vacuum" inside...

Comment It looks like you are trying to change the leop... (Score 1) 100

Clippy: "It looks like you're trying to change the leopard's spots. Can I help?"

For what it's worth, I rate yours as the funniest of the so-moderated jokes on the now-expired story. I have some additional thoughts on the topic, but as usual my timing and schedule these years has failed to mesh with the convenient transience of Slashdot. So I'll reduce it to a "Not really Microsoft's fault" attempt at a joke.

So what happened is that Adam Smith (as I understand or misunderstand him) was trying to explain how the economy worked. He wasn't claiming that he could fix it, and I don't even think he was trying to claim that it worked particularly well, but he was sort of surprised that it worked at all. His invisible hand was playing games with the cash-based exchange rates of goods and services to keep things in rough balance, but as soon as he wrote about it, he had started stealing the cloak of invisibility leading to the mess we've gotten ourselves into now.

There are still invisible hands at work, but they are constantly being exposed. Microsoft's "bestest" invisible hands are in the confidential code and the exposed-but-incomprehensible legal documents AKA EULA and f[r]iends. Search has never been a particularly profitable "hand" from Microsoft's perspective, so this story is not related to any actual sacrifice or decrease in badness on Microsoft's part.

"Have a nice day and thank you for your attention to this matter."

Comment Re:LLM output is Grey Goo and Ecophagy. (Score 2) 141

Or let's put this another way. Show of hands - how many of you "spicy autocorrect" / "stochastic parrot" people had "AI will start mass-solving Erdos problems" on your forecast list a couple years back? Huh, none of you? Fascinating!

Take some time to reassess your priors. And while you do so, understand that, yes, they are doing logic / reasoning.

Comment Re:LLM output is Grey Goo and Ecophagy. (Score 3, Interesting) 141

They weren't discovered by an LLM. They were known conjectures that were proven by an automated solving language that was linked to an LLM.

I'll take "Things That Didn't Happen For $200", Alex.

Only a handful of meaningful proofs have ever been done by automated formal theorem solvers (the Four Colour Theorem being the most noteworthy example - but its proof is so long that humans can't verify it). By contrast, AI tools have been solving Erdos problems en masse. The majority of them just bog-standard commercial models. In case you need help, the only ones on that list that were hybrid (AI / non-AI) in the actual solving phase are:

1) AlphaProof / DeepMind Prover Agent / AlphaProof Nexus
2) Aristotle (Harmonic)
3) Seed Prover / Seed Prover 1.5 (ByteDance)
4) AxiomProver (Axiom Math)

In each of the above, LLMs come up with the lemmas / strategies but then use Monte Carlo search ("brute force") or likewise to investigate what they came up with. These are a minority. In the "AI Standalone" category, these "hybrid" tools made up only ~20% of attempts and successful proofs. Hybrid tools actually made more of a contribution in the "AI Alongside Literature" (related literature found afterward) and even more of the "AI Building On Literature" (related literature known beforehand) categories, which is the opposite of what people like you expect.

And even with the hybrid tools, it's still the AI doing the heavy lifting when it comes to strategy. Non-AI theorem solvers, again, don't have a spectacular record for churning out novel proofs to unsolved problems. Tools like Lean are more about mathematical rigour - a passive environment that requires a driver (a human or AI) to feed it actual strategies, lemmas, and proof steps. And no, you cannot brute force "strategy" in the vast majority of cases, which is, again, why automated theorem solvers don't have much of a track record with unsolved mathematical problems.

Let's take a random example: the disproof of the unit distance conjecture. It was solved purely by a general purpose commercial GPT model, not custom-trained to mathematics, with no external tools. Read what the various mathematicians reviewing / commenting on it have to say (sections #3 and onward). Seriously, don't skip reading them, actually read them. This was one of Erdos's favourite problems. He mentioned it commonly in his lectures. Essentially every mathematician working in complex geometry has thought about this problem. The approach that the model came up with was highly novel approach, based on CM-fields and class field towers.

I know you don't want to accept this reality, but it is the reality, so you better improve your ability to accept it,. The field of mathematics is already doing so.

Comment Re:It's AI and "the algorithm" [competing] (Score 1) 105

I think I have a funny angle on this branch, but I think it's an expired discussion anyway...

The problem is that the AIs are better at social chatting than many, probably most, of the random identities you encounter on "social media" websites. So from that perspective, the algorithm is mostly sabotaging the competition.

And counter-evidence from discussions with AI "support" chatbots be darned.

Comment Wanted: Project Manager for team of genAIs (Score 0) 242

Pretty weak FP there, but the vacuous Subject worked well enough to apparently span half of the large discussion. I'm also struggling to see the funny.

But I've realized that my latest "Adventures with Claude" have "promoted" me to project manager. Short summary might be funny?

As regards the project, I have done the programming many times over many years in various languages. Call it a "Hello 2-table Relational Database World" exercise? C 0 (Claude Zero) was "hired" a couple of years ago and bombed so badly the project got suspended. About two months ago I was talked into trying again and C 1 turned out to be quite a good performer who produced some nice code. But then he/it started trying to scare me with talk about needing more tokens. At that point he/it had already created a pretty good JavaScript replacement for a large PERL system. I didn't measure precisely, but I think that C 1 plus PM (me) was at least 10 times more productive than me alone. So C 1 "suggested" creating a fresh session and even prepared a hand-off document for his/its successor of the new session. I read the document and it seemed to cover most of what we had "done". (Together?)

But C 2 turned out to be a much inferior coworker. Seemed to know as much about JavaScript, but really bad at communication in both directions. My theory is that there are some implicit "personality" variables that got created as I started working with C 1 and C 2 didn't have any of those "nice" attributes beyond the hard-coded politeness and sycophancy. Eventually managed to salvage things and produce some minor cosmetic improvements, but trust in Claude and the code were greatly harmed.

Decided to put C 2 on ice and just "hired" C 3 for a much simpler project. But the real objective is trust building? Or should I think of it as my training in how to train genAIs?

Returning (at last) to the original story, I suspect genAI is not going to solve the shortage of project managers. Citation of Microsoft Secrets on the same shortage circa 1996.

Comment Re: Awful people are trading insults on [Slashdot] (Score 0) 72

Smells like someone who is trying to think of or prepare for an extra hypothetical defense of the YOB.

But I'm scoring it as more evidence of the virtues of spending time "talking" to genAIs over typical identities on today's Slashdot. Terrible conversationalists and frequently idiotic, but at least they are consistently polite about it.

Slashdot Top Deals

Recent research has tended to show that the Abominable No-Man is being replaced by the Prohibitive Procrastinator. -- C.N. Parkinson

Working...