[ryanMVP]

Reading the whitepaper, the inference provider still has the ability to access the prompt and response plaintext. This scheme does seem to guarantee that plaintext cannot be read for all other parties (e.g. the API router), and that the client's identity is hidden and cannot be associated with their request. Perhaps the precise privacy guarantees and allowances should be summarized in the readme.

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?

[macrael]

Howdy, head of Eng at confident.security here, so excited to see this out there.

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.

[bjackman]

> no one, not even people operating the inference hardware

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.

[macrael]

I'm happy to be careful, you are right we are relying on TEEs and vTPMs as roots of trust here and TEEs have been compromised by attackers with physical access.

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.

[michaelt]

> I came to understand that basically there's no HW that actually considers physical attacks in scope for their threat model?

xbox, playstation, and some smartphone activation locks.

Of course, you may note those products have certain things in common...

[bjackman]

Yeah that's a good point, I don't call that confidential compute though it's a different use case.

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.

[bragr]

All things that were compromised with physical attacks? What are mod chips if not physical attack as a service?

[jon-wood]

I'm not aware of working jailbreaks for either Xbox Series or PS5. Its possible that's just a matter of time, but they've both been out for quite a while now it seems like the console manufacturers have finally worked out how to secure them.

[einsteinx2]

Older firmware versions of PS5 are in fact jailbroken (google ps5 jailbreak and you’ll find a bunch of info). I’m not aware of any for Xbox Series but I think that’s more due to lack of interest and the fact that you can run homebrew in development mode already.

[zeusk]

Nvidia has been investing in confidential compute for inference workloads in cloud - that covers physical ownership/attacks in their thread model.

https://www.nvidia.com/en-us/data-center/solutions/confident...

https://developer.nvidia.com/blog/protecting-sensitive-data-...

[bjackman]

It's likely I'm mistaken about details here but I _think_ tee.fail bypassed this technology and the AT article covers exactly that.

[jiveturkey]

> The privacy guarantee we are making here is that no one, not even people operating the inference hardware, can see your prompts.

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.

[dcliu]

That's a strong claim for not looking into it at all.

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.

[jiveturkey]

I guess you're unaware that Intel TEE does not provide physical protection. Literally out of scope, at least per runZero CEO (which I didn't verify). But anyway, in scope or not, it doesn't succeed at it.

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.

[9dev]

It doesn’t matter either way. Intel is an American company as well, and thus unsuitable as a trust root.

[bangaladore]

A company of what country would you prefer?

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).

[9dev]

A European one. Pulling the kind of tricks the NSA does is considerably harder if you don’t have a secret court with secret orders.

[jiveturkey]

> Intel is an American company

Literally.

[brookst]

If you’re moving the goalposts from tech implementation to political vibes, it’s just more post-fact nabobism.

[9dev]

"SSL added and removed here :-)"

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?

[macrael]

Apple actually attests to signatures of every single binary they install on their machines, before soft booting into a mode where no further executables can be installed: https://security.apple.com/documentation/private-cloud-compu...

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.

[jiveturkey]

> Apple actually attests to signatures

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.

[brookst]

You should read the PCC paper: https://security.apple.com/blog/private-cloud-compute/

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.

[astrange]

> OTOH they gave all the data to China for Chinese users, so YMMV

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.

[ryanMVP]

Thanks for the reply! By "inference provider" I meant someone operating a ComputeNode. I initially skimmed the paper, but I've now read more closely and see that we're trying to get guarantees that even a malicious operator is unable to e.g. exfiltrate prompt plaintext.

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.

[macrael]

Great questions!

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

[Terretta]

> the inference provider still has the ability to access the prompt and response plaintext

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.

[michaelt]

> 3. There is no mechanism for a cloud service provider employee to access customer content stored on instance storage and encrypted EBS volumes.

Are you telling me customer services can't reset a customer's forgotten console login password?

[Terretta]

In these systems secured to this degree, yes.

[jiggawatts]

My logic is that these "confidential compute" problems suffer from some of the same issues as "immutable storage in blockchain".

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.)

[avianlyric]

This is hardly a new problem that only appears in the cloud. Any bank that offers a private secure storage facility I.e. a safety deposit box, or anyone that offers a PO Box service is also exposed to the same risk.

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.

[bangaladore]

> 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.

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.

[saithound]

You're simply wrong about this. "I don't know the key" is not legal defense even against hosting an encrypted blob of copyright infringing contemt, much less an encrypted blob of illegal pornography.

[bangaladore]

If this were the case nobody would ever offer file hosting services (eg. Google Drive). Do you have any case history to show any company getting prosecuted for unknowingly hosting encrypted blobs of illegal material?

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"

[saithound]

I think you misunderstand jiggawatt, who wasn't talking about unknowingly hosting illegal material.

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]

[1] https://ansuz.sooke.bc.ca/entry/23

[brookst]

You’ve hit on exactly why Apple proposed that client-side, pre-upload CSAM detection which everyone freaked out about.

Instead, iCloud, Google Drive, and similar all rely on being able to hash content post-upload for exactly that reason.

[sublimefire]

Yes but at the end of the day you need to trust the cloud provider tools which expands the trust boundary from just hardware root of trust. Who is to guarantee they will not create a malicious tool update and push it then retract it? It is nowhere captured and you cannot prove it.

[avianlyric]

You can detect and prove it because the hardware attestation signature will change.

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.

[7e]

At the end if the day, Nitro Enclaves are still “trust Amazon”, which is a poor guarantee. NVIDIA+AMD offers hardware backed enclave features for their GPUs which is the superior solution here.

[iancarroll]

Aren’t they both hardware backed, just changing the X in “trust X”?

[Terretta]

You think Nitro Enclaves aren't hardware backed?

[almostgotcaught]

be sure to let us know when you can run eg nginx on a GPU in said enclave.

[amelius]

> Folks may underestimate the difficulty of providing compute that the provider “cannot”* access to reveal even at gunpoint.

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.

[astrange]

That's what non-targetability is for.

[immibis]

It's probably illegal for a business to take anonymous cryptocurrency payments in the EU. Businesses are allowed to take traceable payments only, or else it's money laundering.

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...

[anon721656321]

at that point, it seems easier to run a slightly worse model locally. (or on a rented server)

[rimeice]

Which is apples own approach until the compute requirements need them to run some compute on cloud.

[bigyabai]

Just a shame they spent so long skimping on iPhone memory. The tail-end of support for 4gb and 6gb handsets is going to push that compute barrier pretty low.

[brookst]

Eh, maybe a bit, but those era devices also have much lower memory bandwidth. I suspect that the utility of client models will rule out those devices for other reasons than memory.

[bigyabai]

> much lower memory bandwidth

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.

[rasengan]

We are introducing Verifiably Private AI [1] which actually solves all of the issues you mention. Everything across the entire chain is verifiably private (or in other words, transparent to the user in such a way they can verify what is running across the entire architecture).

[1] https://ai.vp.net/

[jmort]

It should be able to support connecting via an OpenPCC client, then!

[poly2it]

Whitepaper?