Bringing FHE onchain @fhenix; Blockchain Privacy pioneer (Enigma, Secret Network); @MIT PhD in Cryptography
Joined July 2010
- Tweets 5,417
- Following 476
- Followers 33,823
- Likes 10,108
In parenting - there's what you say and what you do. The latter is much more important than the first.
If governments are supposed to be the responsible adult, but they're taking on more and more debt, there's no wonder us plebs follow suit.
We now live in a society where more debt is the “solution” to everything:
Can’t afford basic necessities? Take on more debt.
Can’t pay your tuition? Take on more debt.
Can’t afford to buy a home? Take on more debt.
Can’t keep inflation under control? Take on more debt.
Can’t run the government efficiently? Take on more debt.
There’s a reason why consumer sentiment is now below 2008 levels.
Our society is drowning in debt.
If FHE provides the same privacy guarantees as TEEs.
Then zk provides the same integrity guarantees as TEEs.
There is no way to believe the first without also admitting the second.
Both are false, clearly.
Difference between TEE, ZK, MPC and FHE -- summing up:
(I know I wrote a lot about it over the past 2 weeks. For those who survived, you get a prize: fewer words about it.)
TEEs -
Promise to solve integrity and privacy but have a single point of failure: physical keys that can be extracted by malicious operators (Lazarus group and friends).
Should never be used in truly decentralized blockchain infra.
(TEEs can be used in permissioned settings, usually as security theater, or for end-user devices and small sums).
ZK -
The most mature and battle tested of the technologies surveyed here.
It has great smart contract languages – Cairo, has been used for years in Zcash and by StarkWare (and others) and gives individual privacy and integrity at immense scale.
MPC -
Preserves privacy in multi-party settings (like running an auction for an NFT) but isn’t as mature a technology.
It’s difficult to use because its interactive (all users have to send messages back and forth) and doesn’t have a smart contract language like Cairo.
FHE -
Gives privacy guarantees similar to TEE, but actually delivers on that without any leakholes. Like MPC is newer technology, cannot support general languages like Cairo, and for most use cases in blockchain requires the assistance of ZK technology for proving integrity.
Links to these posts in the comments.
Thanks for reading and of course, if you have a good way to explain these concepts I encourage you to share it here in the comments. There's no such thing as too much education.
THE END
This is so fucking messed up.
* Just this year @POTUS pardoned a dev sent to prison for writing code. Have we learned nothing?
* Lack of privacy in crypto is a bug. The fact that devs are trying to fix it should be praised, not punished.
* I will be the first to say we need compliant privacy -- but products like Samourai wallet were very early and very limited in functionality. There are better ways to do things today (with any form of programmable privacy -- FHE, MPC, or even by adding additional wrappers to zkMixers, like Privacy Pools does).
Wake up USA, you're going backwards.
🚨KEONNE RODRIGUEZ SENTENCED TO 5 YEARS IN PRISON, $250,000 FINE, 3 YEARS OF SUPERVISED RELEASE
The Samourai Wallet developer will surrender to custody on December 19th. Rodriguez has the option to appeal, Judge Cote states.
Full story👇
therage.co/keonne-rodriguez-…
Encryption Days by @fhenix were consistently the most well-attended privacy event in the last two years.
Join us on Nov 19 for Encryption Day @ Devconnect version: luma.com/e9ces3lr.
While our events were always full, they were never packed. We always consistently had hundreds of people attend - mostly builders and researchers.
But given that privacy is booming, I have a feeling this time will be different and the day will be super packed.
And this scares the shit out of introvert me with my stage fright.
I think I saw the first major player to implode this cycle. Still unverified, but serious FTX/Celsius vibes..
It will never cease to amaze me how multi-billion dollar companies in this space YOLO like a gambling addict in Vegas.
Can we as an industry agree on what's the best name for secure computation in blockchains is? For such a fundemental piece of the puzzle, it's insane we don't have a clear-cut term. Is it:
* Compute is the keyword: secure -OR- private -OR- confidential -OR- encrypted computation
* Smart contracts are the keyword: confidential/private/encrypted smart contracts
* Something novel: Either programmable privacy or verifiable privacy
* Other?
.@AvishaiY @ittaia @luhelminger need your feedback here as well.
I don't want this to be project-specific, and we can definitely treat this as a first approximation we can change/update/debate.
It's been a gap for us to communicate the privacy guarantees simply and consistently - I'm certain it also comes up in your conversations.
I’d love to see @ethereum go beyond just pushing ZK for privacy and discuss other technologies like FHE/MPC/TEEs for privacy.
Here are two talks I gave about private onchain compute 7-8 years ago (Devcon 3 and 4).
piped.video/CtyEOQ7zckE?si=R6GO…
piped.video/T47CneAXJVg?si=a7LR…
0/ Privacy has always been at the heart of the Ethereum ecosystem.
Here are 8 privacy talks from the last 10 years of @EFDevcon events.
Privacy is normal.
(Guy Zyskind*) retweeted
Wow, major respect for @EliBenSasson, you are a legend, but 90% of this post is bullshit.
The sad part is that's not ignorance (he is a genius), but just narrative games.
Wish we'd be able to focus on actually building the privacy space, rather than 💩ing on eachother
Privacy matters --
Difference between TEE, ZK, MPC and FHE - Last one!
Today we explain: FHE
FHE - Fully Homomorphic Encryption is a very powerful cryptographic privacy paradigm.
It solves confidentiality, but not integrity. In this sense it's complimentary to ZK, and in an ideal world you’d want to use the two of them combined.
FHE gives you the same privacy guarantees as a TEE:
- You send encrypted information to an FHE circuit
- It performs a computation on your data. It does it without knowing anything about your data.
- Then it sends you the result in encrypted form, for you to decrypt.
This is useful in settings like storing your data on the cloud, and asking the cloud provider to compute on your encrypted data while it remains encrypted.
The problems with FHE are several:
- First, like the case of MPC, you can’t really use modern programming languages like Cairo for writing smart contracts. You have to work with circuits, which are more limited.
- The second problem is that FHE doesn’t deliver integrity, so if you ask the blockchain to run FHE you will need *all* nodes to re-execute that computation, which is very costly.
- The third issue is that of maturity. Like MPC (and unlike ZK), so far there’s been limited experience and usage of it, but I hope this will change with time.
The END
(Guy Zyskind*) retweeted
With full respect for your experience and knowledge, I think your recent post was misleading and quite dishonest.
> you can’t really use modern programming languages like Cairo for writing smart contracts
Calling Solidity, which is supported by nearly all FHE solutions, a non-modern language, and at the same time calling Cairo, which is used only within a single ecosystem, modern, feels inconsistent and sounds more like a marketing move than a genuine technical view.
> FHE gives you the same privacy guarantees as a TEE
Having worked with both, I can say that claiming FHE, an LWE-based quantum-secure scheme, provides the same guarantees as secure hardware goes against every cryptographic principle I’ve encountered.
To clarify, I agree with part of your last point. FHE is far more mature than it has ever been, but there are still crucial steps ahead before it reaches full maturity. Compared to ZK, which had about a decade’s head start, FHE is on a very promising track.
Privacy matters --
Difference between TEE, ZK, MPC and FHE - Last one!
Today we explain: FHE
FHE - Fully Homomorphic Encryption is a very powerful cryptographic privacy paradigm.
It solves confidentiality, but not integrity. In this sense it's complimentary to ZK, and in an ideal world you’d want to use the two of them combined.
FHE gives you the same privacy guarantees as a TEE:
- You send encrypted information to an FHE circuit
- It performs a computation on your data. It does it without knowing anything about your data.
- Then it sends you the result in encrypted form, for you to decrypt.
This is useful in settings like storing your data on the cloud, and asking the cloud provider to compute on your encrypted data while it remains encrypted.
The problems with FHE are several:
- First, like the case of MPC, you can’t really use modern programming languages like Cairo for writing smart contracts. You have to work with circuits, which are more limited.
- The second problem is that FHE doesn’t deliver integrity, so if you ask the blockchain to run FHE you will need *all* nodes to re-execute that computation, which is very costly.
- The third issue is that of maturity. Like MPC (and unlike ZK), so far there’s been limited experience and usage of it, but I hope this will change with time.
The END
Saying that zk devex/programmability > FHE is absolutely dishonest.
No one writes FHE circuits. You write programs. In fact, you write Solidity programs. We've solved that @fhenix long ago, and we're not the only ones.
Also, saying that giving developers a new DSL language that can only be used in a single ecosystem (Starkware) is better devex is dishonest.
Writing FHE smart contracts is super easy. You don't need to deal with circuits at all. Just import the FHE library and write Solidity - github.com/FhenixProtocol/en…
Privacy matters --
Difference between TEE, ZK, MPC and FHE - Last one!
Today we explain: FHE
FHE - Fully Homomorphic Encryption is a very powerful cryptographic privacy paradigm.
It solves confidentiality, but not integrity. In this sense it's complimentary to ZK, and in an ideal world you’d want to use the two of them combined.
FHE gives you the same privacy guarantees as a TEE:
- You send encrypted information to an FHE circuit
- It performs a computation on your data. It does it without knowing anything about your data.
- Then it sends you the result in encrypted form, for you to decrypt.
This is useful in settings like storing your data on the cloud, and asking the cloud provider to compute on your encrypted data while it remains encrypted.
The problems with FHE are several:
- First, like the case of MPC, you can’t really use modern programming languages like Cairo for writing smart contracts. You have to work with circuits, which are more limited.
- The second problem is that FHE doesn’t deliver integrity, so if you ask the blockchain to run FHE you will need *all* nodes to re-execute that computation, which is very costly.
- The third issue is that of maturity. Like MPC (and unlike ZK), so far there’s been limited experience and usage of it, but I hope this will change with time.
The END
Infrastructure still has a place, but it’s neither as an L1 or an L2.
Existing chains need expansion packs.
Deciding to move away from the L1/L2 meta was a hard choice - but it was the smart product choice.
Let’s talk Monad.
Thanks for the stimmy check. Seems like it will be 0 before we even see it.
I already predicted the playbook awhile back when I talked about Berachain.
Alt-L1s are completely pointless. We already have Ethereum and Solana. Any new alt-L1 raising 9 figures from VCs is to extract value from retail.
The framework is established, we need more killer apps, not more chains.
Someone got FRIed
An exciting update from myself and @benediamond (eprint.iacr.org/2025/2010). We show that the 𝘶𝘱-𝘵𝘰-𝘤𝘢𝘱𝘢𝘤𝘪𝘵𝘺 proximity gaps conjecture is 𝗳𝗮𝗹𝘀𝗲. More precisely, given any pair c, d we construct codes whose error grows faster than nᶜ / (q ⋅ (ρ η)ᵈ).
