With that in mind, does this scheme offer any advantage over the much simpler setup of a user sending an inference request:
- directly to an inference provider (no API router middleman)
- that accepts anonymous crypto payments (I believe such things exist)
- using a VPN to mask their IP?
I'm not sure I understand what you mean by inference provider here? The inference workload is not shipped off the compute node once it's been decrypted to e.g. OpenAI, it's running directly on the compute machine on open source models loaded there. Those machines are cryptographically attesting to the software they are running. Proving, ultimately, that there is no software that is logging sensitive info off the machine, and the machine is locked down, no SSH access.
This is how Apple's PCC does it as well, clients of the system will not even send requests to compute nodes that aren't making these promises, and you can audit the code running on those compute machines to check that they aren't doing anything nefarious.
The privacy guarantee we are making here is that no one, not even people operating the inference hardware, can see your prompts.
You need to be careful with these claims IMO. I am not involved directly in CoCo so my understanding lacks nuance but after https://tee.fail I came to understand that basically there's no HW that actually considers physical attacks in scope for their threat model?
The Ars Technica coverage of that publication has some pretty yikes contrasts between quotes from people making claims like yours, and the actual reality of the hardware features.
https://arstechnica.com/security/2025/10/new-physical-attack...
My current understanding of the guarantees here is:
- even if you completely pwn the inference operator, steal all root keys etc, you can't steal their customers' data as a remote attacker
- as a small cabal of arbitrarily privileged employees of the operator, you can't steal the customers' data without a very high risk of getting caught
- BUT, if the operator systematically conspires to steal the customers' data, they can. If the state wants the data and is willing to spend money on getting it, it's theirs.
This is actually part of why we think it's so important to have the non-targetability part of the security stack as well, so that even if someone where to physically compromise some machines at a cloud provider, there would be no way for them to reliably route a target's requests to that machine.
xbox, playstation, and some smartphone activation locks.
Of course, you may note those products have certain things in common...
CoCo = protecting consumer data from the industry. DRM = protecting industry bullshit from the consumer.
TBF my understanding is that in the DRM usecases they achieve actual security by squeezing the TCB into a single die. And I think if anyone tries, they generally still always get pwned by physical attackers even though it's supposedly in scope for the threat model.
https://www.nvidia.com/en-us/data-center/solutions/confident...
https://developer.nvidia.com/blog/protecting-sensitive-data-...
that cannot be met, period. your asssumptions around physical protections are invalid or at least incorrect. It works for Apple (well enough) because of the high trust we place in their own physical controls, and market incentive to protect that at all costs.
> This is how Apple's PCC does it as well [...] and you can audit the code running on those compute machines to check that they aren't doing anything nefarious.
just based on my recollection, and I'm not going to have a new look at it to validate what I'm saying here, but with PCC, no you can't actually do that. With PCC you do get an attestation, but there isn't actually a "confidential compute" aspect where that attestation (that you can trust) proves that is what is running. You have to trust Apple at that lowest layer of the "attestation trust chain".
I feel like with your bold misunderstandings you are really believing your own hype. Apple can do that, sure, but a new challenger cannot. And I mean your web page doesn't even have an "about us" section.
From a brief glance at the white paper it looks like they are using TEE, which would mean that the root of trust is the hardware chip vendor (e.g. Intel). Then, it is possible for confidentiality guarantees to work if you can trust the vendor of the software that is running. That's the whole purpose of TEE.
And I mean I get it. As a not-hardware-manufacturer, they have to have a root of trust they build upon. I gather that no one undertakes something like this without very, very, very high competence and that their part of the stack _is_ secure. But it's built on sand.
I mean it's fine. Everything around us is built that way. Who among us uses a Raptor Talus II and has x-ray'd the PCB? The difference is they are making an overly strong claim.
Everyone likes to dunk on the US, but I doubt you could provide a single example of a country that is certainly a better alternative (to be clear I believe many of the west up in the same boat).
Literally.
It’s not about vibes, but clear proof of a strategy to undermine global information security. Is anyone suppose to believe they don’t do that anymore?
We don't _quite_ have the funding to build out our own custom OS to match that level of attestation, so we settled for attesting to a hash of every file on the booted VM instead.
But (based on light reading, forgive errors) the only way to attest them is to ask _Apple_! It reminds me what i call e2e2e encryption. iMessage is secure e2e but you have to trust that Apple is sending you the correct keys. (There's some recent update, maybe 1-2 years old, where you can verify the other party's keys in person I think? But it's closed software, you _still_ have to trust that what you're being shown is something that isn't a coordinated deception.)
Apple claims to operate the infrastructure securely, and while I believe they would never destroy their business by not operating as rigorously as they claim, OTOH they gave all the data to China for Chinese users, so YMMV. And their OS spams me with ads for their services. I absolutely hate that.
Again, anyway, I am comfortable putting my trust in Apple. My data aren't state secrets. But I wouldn't be putting my trust in random cloud operator based on your known-invalid claim of physical protection. Not if the whole point is to protect against an untrustworthy operator. I would much sooner trust a nitro enclave.
You are not in fact trusting Apple at all. You are trusting some limited number of independent security researchers, which is not perfect, but the system is very carefully designed to give Apple themselves no avenue to exploit without detection.
This is true for the same reason that American data is in the US. China is frequently a normal and competent country and has data privacy laws too.
Despite recent news of vulnerabilities, I do think that hardware-root-of-trust will eventually be a great tool for verifiable security.
A couple follow-up questions:
1. For the ComputeNode to be verifiable by the client, does this require that the operator makes all source code running on the machine publicly available?
2. After a client validates a ComputeNode's attestation bundle and sends an encrypted prompt, is the client guaranteed that only the ComputeNode running in its attested state can decrypt the prompt? Section 2.5.5 of the whitepaper mentions expiring old attestation bundles, so I wonder if this is to protect against a malicious operator presenting an attestation bundle that doesn't match what's actually running on the ComputeNode.
1. The mechanics of the protocol are that a client will check that the software attested to has been released on a transparency log. dm-verity is what enforces that the hashes of the booted filesystem on the compute node match what was built and so those hashes are what are put on the transparency log, with a link to the deployed image that matches them. The point of the transparency log is that anyone could then go inspect the code related to that release to confirm that it isn't maliciously logging. So if you don't publish the code for your compute nodes then the fact of it being on the log isn't really useful.
So I think the answer is yes, to be compliant with OpenPCC you would need to publish the code for your compute nodes, though the client can't actually technically check that for you.
2. Absolutely yes. The client encrypts its prompt to a public key specific to a single compute node (well, technically it will encrypt the prompt N times for N specific compute nodes) where the private half of that key is only resident in the vTPM, the machine itself has no access to it. If the machine were swapped or rebooted for another one, it would be impossible for that computer to decrypt the prompt. The fact that the private key is in the vTPM is part of the attestation bundle, so you can't fake it
Folks may underestimate the difficulty of providing compute that the provider “cannot”* access to reveal even at gunpoint.
BYOK does cover most of it, but oh look, you brought me and my code your key, thanks… Apple's approach, and certain other systems such as AWS's Nitro Enclaves, aim at this last step of the problem:
- https://security.apple.com/documentation/private-cloud-compu...
- https://aws.amazon.com/confidential-computing/
NCC Group verified AWS's approach and found:
1. There is no mechanism for a cloud service provider employee to log in to the underlying host.
2. No administrative API can access customer content on the underlying host.
3. There is no mechanism for a cloud service provider employee to access customer content stored on instance storage and encrypted EBS volumes.
4. There is no mechanism for a cloud service provider employee to access encrypted data transmitted over the network.
5. Access to administrative APIs always requires authentication and authorization.
6. Access to administrative APIs is always logged.
7. Hosts can only run tested and signed software that is deployed by an authenticated and authorized deployment service. No cloud service provider employee can deploy code directly onto hosts.
- https://aws.amazon.com/blogs/compute/aws-nitro-system-gets-i...
Points 1 and 2 are more unusual than 3 - 7.
Folks who enjoy taking things apart to understand them can hack at Apple's here:
https://security.apple.com/blog/pcc-security-research/
* Except by, say, withdrawing the system (see Apple in UK) so users have to use something less secure, observably changing the system, or other transparency trippers.
Are you telling me customer services can't reset a customer's forgotten console login password?
I.e.: If the security/privacy guarantees really are as advertised, then ipso facto someone could store child porn in the system and the provider couldn't detect this.
Then by extension, any truly private system is exposing themselves to significant business, legal, and moral risk of being tarred and feathered along with the pedos that used their system.
It's a real issue, and has come up regularly with blockchain based data storage. If you make it "cencorship proof", the by definition you can't scrub it of illegal data!
Similarly, if cloud providers allow truly private data hosting, then they're exposing themselves to the risk of hosting data that is being stored with that level of privacy guarantees precisely because it is so very, very illegal.
(Or substitute: Stolen state secrets that will have the government come down on you like a ton of bricks. Stolen intellectual property. Blackmail information on humourless billionaires. Illegal gambling sites. Nuclear weapons designs. So on, and so forth.)
But both of these services exist, and have existed for hundreds of years, and don’t require service providers to go snooping though their customer’s possessions or communications.
But what they would be storing in this case is not illegal content. Straight up. Encrypted bits without a key are meaningless.
There is nothing stopping a criminal from uploading illegal content to Google drive as an encrypted blob. There's nothing Google can do about it, and there is no legal repercussion (to my knowledge) of holding such a blob.
Obviously if they have to ability to know material is illegal, that's a problem.
And exactly what algorithm can you provide to me that takes an encrypted blob as an input and returns whether it is not illegal material. Clearly that doesn't exist, so your point makes zero sense.
You may be conflating "I forgot the key" vs "I've never been provided the key"
We're talking about knowingly hosting encrypted illegal material without knowing the key. This is unambiguously illegal whether or not you ever knew the key.
If the police show up and tell you that your site has an encrypted zip file containing illegal porn, of course they can instruct you to stop hosting it, and hold you liable if you refuse to follow those instructions.
They're not going to give you the decryption key to check for yourself, and it'd not even be legal for them to do so.
Jiggawatt is saying that if you have a truly uncensorable system, it's impossible to comply with the police instructions to selectively remove the illegal material, and so the whole thing becomes illegal.
> And exactly what algorithm can you provide to me that takes an encrypted blob as an input and returns whether it is not illegal material. Clearly that doesn't exist, so your point makes zero sense.
This on the other hand, tells me you don't know much about how legal systems work. I recommend you start with the essay "What color are your bits?" [1]
My point all along, is that Google is not liable for someone uploading previously encrypted blobs of illegal content to Google Drive. And even more so, Google isn't liable if someone uploads illegal content to Google Drive that isn't encrypted. Google simply needs to remove it and follow the correct processes if reported / detected.
Could you make an argument for either that theoretically they could be? Sure. But in reality, no, they are not liable.
This is law due to Section 230:
> Section 230 of the Communications Act of 1934, enacted as part of the Communications Decency Act of 1996, provides limited federal immunity to providers and users of interactive computer services. The statute generally precludes providers and users from being held liable—that is, legally responsible—for information provided by another person, but does not prevent them from being held legally responsible for information that they have developed or for activities unrelated to third-party content. Courts have interpreted Section 230 to foreclose a wide variety of lawsuits and to preempt laws that would make providers and users liable for third-party content. For example, the law has been applied to protect online service providers like social media companies from lawsuits based on their decisions to transmit or take down user-generated content.
https://www.congress.gov/crs-product/R46751 https://www.eff.org/issues/cda230
Also, the blockchain problem already exists. I've linked some commentary about it.
https://ethereum.stackexchange.com/questions/94558/what-prev...
Instead, iCloud, Google Drive, and similar all rely on being able to hash content post-upload for exactly that reason.
You might not know what change was made, or have any prior warning of the change. But you will be able to detect it happening. Which means an operator only gets to play that card once, after which nobody will trust them again.
It's even harder to do this plus the hard requirement of giving the NSA access.
Or alternatively, give the user a verifiable guarantee that nobody has access.
With the caveat that it's not clear what precisely is illegal about these payments and to what level it's illegal. It might be that a business isn't allowed to have any at all, or isn't allowed to use them for business, or can use them for business but can't exchange them for normal currency, or can do all that but has to check their customer's passport and fill out reams of paperwork.
https://bitcoinblog.de/2025/05/05/eu-to-ban-trading-of-priva...
Not really? The A11 Bionic chip that shipped with the iPhone X has 3gb of 30gb/s memory. That's plenty fast for small LLMs if they'll fit in memory, it's only ~1/3rd of the M1's memory speed and it only gets faster on the LPDDR5 handsets.
A big part of Apple's chip design philosophy was investing in memory controller hardware to take advantage of the iOS runtime better. They just didn't foresee any technologies beside GC that could potentially inflate memory consumption.
Service: https://www.privatemode.ai/ Code: https://github.com/edgelesssys/privatemode-public
It's better than nothing, I guess...
But if you placed the server at the NSA, and said "there is something on here that you really want, it's currently powered on and connected to the network, and the user is accessing it via ssh", it seems relatively straightforward for them to intercept and access.
2) These attacks are actually worse than what I am pretty sure you are assuming (and so where I started my response), as you actually just need one hacked server and then you can simulate working servers on other hardware that isn't hacked by either stealing an attested key or stealing the attestation key itself. You often wouldn't even then need to have the hacked server anymore.
> targeting users should require a wide attack that’s likely to be detected
Regardless, Apple's attacker here doesn't sound like Apple: the "wide attack that's likely to be detected" is going to be detected by them. We even seemingly have to trust them that this magic hardware has the properties they claim it does.
This is way worse than most of these schemes, as if I run one of these on Intel hardware, you inherently are working with multiple parties (me and Intel).
That we trust Apple to not be lying about the entire scheme so they can see the data they are claiming not to be able to see is thereby doing the heavy lifting.
we are working on a challenge which is somewhat like a homomorphic encryption problem - I'm wondering if OpenPCC could help in some way? :
When developing websites/apps, developers generally use logs to debug production issues. However with wearables, logs can be privacy issue: imagine some AR glasses logging visual data (like someone's face). Would OpenPCC help to extract/clean/anonymize this sort of data for developers to help with their debugging?
If it's possible to anonymize on the wearable, that would be simpler.
The challenge is what does the anonymizer "do" to be perfect?
As an aside, IMO homomorphic encryption (still) isn't ready...
[1] https://techcommunity.microsoft.com/blog/azureconfidentialco...
Edit : reminds me of federated learning and FlowerLLM (training only AFAIR, not inference), like... yes, nice, I ALWAYS applaud any way to disentangle from proprieaty software and wall gardens... but like what for? What actual usage?
Edit on that too : makes me think of OpenAI Whisper as a service via /e/OS and supposedly anonymous proxying (by mixing), namely running STT remotely. That would be an actual potential usage... but IMHO that's low end enough to be run locally. So I'm still looking for an application here.
- useful thing (according to someone specific requirements, maybe hallucinations are OK, maybe not) that
- needs privacy (for example generating code that will be open source probably does not need that)
- can't be run locally
- can be trusted to actually process as said it does
The key is that these are all things I can verify without much difficulty: read over the script, spot-check the analysis, look at the claimed connection between A and B and see if it's real. And I don't really care about style, quality, maintainability.
You certainly can run this locally, but anything that will fit into reasonable local hardware won't be as good.
I don't need to trust it to process as it says it does, because I'm verifying the output.
And as far as I'm concerned, "needs privacy" is always true. I don't care if the code will be open source. I don't care if it's analyzing existing code that's already open source. Other people have no business seeing what I'm doing unless I explicitly allow it. In any case, I work on a lot of proprietary code as well, and my employer would be most displeased if I were exposing it to others.
> Gimme an actual example instead of downvoting, help me learn.
Basically you asked a bunch of people on a privacy minded forum, why should they be allowed to encrypt their data? What are you (they) hiding!? Are you a spammer???
Apple is beloved for their stance on privacy, and you basically called everyone who thinks that's more than marketing, a spammer. And before you start arguing no you didn't, it doesn't matter that you didn't, what matters is that that's how your comment made people feel. You can say they're the stupid ones because that's not what you wrote, but if you're genuinely asking for feedback about the downvotes, there you are.
You seriously can't imagine any reason to want to use an LLM privately other than to use it to write spam bots and to spam people? At the very least expand your scope past spamming to, like, also using it to write ransomware.
The proprietary models that can't be run locally are SOTA and local models, even if they can come close, simply aren't what people want.
I specifically like the privacy aspect (even though honestly I think most people in this forum claim they do and yet they rely on BigTech with which they their data, so IMHO most people on HN are not as demanding as you describe) the question is precisely about what to run.
FWIW I specifically keep a page on self hosting AI (you can check if you want to see if it's real at https://fabien.benetou.fr/Content/SelfHostingArtificialIntel... ) so again, the privacy aspect is crucial to me.
The question is ... what to actually run. What can't be run locally that would be useful. What model for which tasks. For example coding (which I don't think models are good enough for) typically would NOT need this because one wouldn't share any PII over there, hopefully would even instead publish the resulting code as open source.
So my provocation about spammer is... because they often ARE actual users of LLMs. They use LLM for their language capabilities, namely craft a message that is always slightly different yet convey (roughly) the same meaning (the scam) to avoid detection. A random person though using LLM might NOT be OK with hallucinations when they use it for their own private journal or chat.
So... what for?
Edit: I did use Apple for years, recommending by someone at Mozilla, and I moved away from them earlier this year precisely because even though they are better than others, e.g Google, IMHO it's just not good enough for me. No intermediary is better than one with closed source.
It's an LLM. The user can ask it anything they can think of! But then to to "only spammers could eke out anything from them".
What do users ask OpenAI about? My boyfriend/girlfriend/wife/mistress cheated on me, what should I do? My mom is... My dad is... My friends are.. I have this problem with this company and I want to sue them... I have this weird lump on my foot... More nefariously, I'm sure someone out there asking "how do I make cocaine" is seriously considering it, and not just testing the machine. I want to talk to somebody about 1994, the TV series from 2019. I want to write a fiction book about the near future but one where I won the lottery or I grew up rich or I was Harry Potter or a murder mystery or utopian sci-fi or dystopian sci-fi or an alt-history where there are still dinousaurs or or or.
I don't know if it's a failure of your imagination, or if mine is overactive, but making a venn diagram of all the world's humans broken down into spammers and not spammers, and then placing the circle users of local LLMs inside of spammers, and there's no one else, just seems a bit reductive.
I realize this is just bad branding by apple but it's still hella confusing.
As for the matter at hand, I assume that such implementation is validated against which specification exactly?
Gladly to be educated in my clueless.
Apparently some of us actually have a clue.
To make a full implementation of a Apple product, the specification for that Apple product must exist in some form.
I think it's more fair to say it implements the same idea but it is not an opensource implementation of Apple's Private Compute Cloud the way e.g. minio is an implementation of S3, so the HN title is misleading.