LibraryConfidential compute2015Design paperCorpus record
Enigma: Decentralized Computation Platform with Guaranteed Privacy
Secret Network. Guy Zyskind, Oz Nathan, Alex Pentland.
The 2015 Enigma paper from MIT: private contracts executed over secret-shared data, so nodes compute without seeing the raw inputs. Secret Network is a later project in this lineage, using different machinery. This page is about the Enigma paper, which the historic library does not hold.
Enigma's paper splits data so that a set of nodes can compute on it together without any one node seeing the underlying secret, using secret sharing rather than a single trusted box.
- Evidence
- Primary paper
- Re-measured
- No
- Assumptions
- 4
- Records linked
- 5
Code
Source, repository, and what is actually hosted
- Enigma paperhttps://enigma.co/enigma_full.pdf
- The source URL is not a GitHub repository. No code remote is invented.
- Full text is not mirrored. Public availability is not a licence to copy the file onto this site.
- Add to a reading list
01 Claim ledger
What the paper is allowed to say
Each row is a sentence already in the study. The status is the same on every row: a model claim, not a live measurement.
Claim 01 · Paper model
The computation needs the data, and the operator must not have it
That is the problem. Encryption at rest does not solve it, because the computer that runs the job usually decrypts. Enigma's answer is to avoid a single computer that can decrypt.
Claim 02 · Paper model
Secret sharing splits the input
Each node holds a share. A threshold of shares can reconstruct. Fewer cannot. The computation is designed to run on shares.
Claim 03 · Paper model
A public chain is the coordinator, not the data store
The paper uses a blockchain for control, identity and settlement. The private data stays off it, in the shared computation layer.
Claim 04 · Paper model
Contracts can call private computation
The design goal is a public contract that triggers a private function and receives only the result it is supposed to see. Specifying that interface is as important as the cryptography.
Claim 05 · Paper model
Multi-party computation has a performance cost the paper does not wish away
Shares, rounds of communication, and a threshold assumption are the price of not trusting one box. A deployment that quietly uses one server has left the paper.
02 Three cuts
Observation, model, falsifier
A desk does not stop at the summary. Each claim is cut three ways, using only this study's own assumptions and checks. Nothing here is a new figure.
01 The computation needs the data, and the operator must not have it
Observation
What the study says
That is the problem. Encryption at rest does not solve it, because the computer that runs the job usually decrypts. Enigma's answer is to avoid a single computer that can decrypt.
Model
What has to hold
Nodes stay online for the rounds the protocol needs. Multi-party computation does not love absentees.
Falsifier
What would retire it
What is the collusion threshold, in numbers?
02 Secret sharing splits the input
Observation
What the study says
Each node holds a share. A threshold of shares can reconstruct. Fewer cannot. The computation is designed to run on shares.
Model
What has to hold
Fewer than the threshold of nodes collude or are compromised.
Falsifier
What would retire it
Is the live system still secret sharing, or has it moved to enclaves or proofs?
03 A public chain is the coordinator, not the data store
Observation
What the study says
The paper uses a blockchain for control, identity and settlement. The private data stays off it, in the shared computation layer.
Model
What has to hold
Nodes stay online for the rounds the protocol needs. Multi-party computation does not love absentees.
Falsifier
What would retire it
What metadata remains public on the coordinating chain?
04 Contracts can call private computation
Observation
What the study says
The design goal is a public contract that triggers a private function and receives only the result it is supposed to see. Specifying that interface is as important as the cryptography.
Model
What has to hold
The function being evaluated was correctly turned into a share-wise protocol. A bad circuit is a bad result the cryptography will faithfully compute.
Falsifier
What would retire it
What metadata remains public on the coordinating chain?
05 Multi-party computation has a performance cost the paper does not wish away
Observation
What the study says
Shares, rounds of communication, and a threshold assumption are the price of not trusting one box. A deployment that quietly uses one server has left the paper.
Model
What has to hold
Nodes stay online for the rounds the protocol needs. Multi-party computation does not love absentees.
Falsifier
What would retire it
What is the collusion threshold, in numbers?
03 Sequence
One action, as an operating tape
- 01A user secret-shares an input among the computation nodes. No single node receives the plaintext.
- 02A public contract records the request and the identity of the function to run.
- 03Nodes run the agreed protocol on their shares, exchanging messages the protocol requires.
- 04The result, and only the result, is returned to the contract or the user. Intermediate shares are not published.
- 05If fewer than the threshold of nodes are honest, the privacy or the correctness claim fails in the way the paper's model predicts.
04 Load-bearing
The argument, and where a pitch drops it
The threshold is the trust model
The cut
Secret sharing does not remove trust. It changes 'trust this operator' into 'do not let this many of these operators collude'. A study should say the number. Without the number, the design is a mood.
Why it carries weight
If this cut is skipped, the paper's name is being used without the mechanism that makes the name mean anything.
Where it is dropped
This is not the Secret Network documentation.
Correctness and privacy are different failures
The cut
Nodes can try to learn the secret, or they can try to return the wrong result. A protocol may defend one better than the other. The paper has to be read for both. A marketing line about 'encrypted computation' usually blurs them.
Why it carries weight
If this cut is skipped, the paper's name is being used without the mechanism that makes the name mean anything.
Where it is dropped
This is not the Secret Network documentation.
The chain sees the request
The cut
Metadata, function names, and who called are often public even when inputs are shared. For many businesses the metadata is the secret. The paper's split helps only if the public half has been examined.
Why it carries weight
If this cut is skipped, the paper's name is being used without the mechanism that makes the name mean anything.
Where it is dropped
This is not the Secret Network documentation.
05 Register
What has to be true
Stake and threshold · Not re-measured
Fewer than the threshold of nodes collude or are compromised.
Cryptography · Not re-measured
The function being evaluated was correctly turned into a share-wise protocol. A bad circuit is a bad result the cryptography will faithfully compute.
Network · Not re-measured
Nodes stay online for the rounds the protocol needs. Multi-party computation does not love absentees.
Model · Not re-measured
Users can verify they are talking to the real set of nodes, not a lookalike that asks for plaintext 'as a convenience'.
06 Divergence
What happened after the paper
Enigma the project evolved, and the broader field moved through multi-party computation, hardware enclaves and zero-knowledge proofs as competing answers to private computation. This paper is the secret-sharing answer. It should not be cited as a description of a later product that swapped in a different foundation.
A later client, parameter set, or reward formula is a different object. Cite this paper for the mechanism. Cite a primary release for the network. This desk has not re-run the proof.
07 Pre-mortem
What to check before you use the idea
- 0 of 4 marked on this browser. A mark is a reading note, not a pass, a rating, or a recommendation.
08 Anatomy
The paper, in the order a builder needs
The problem it names
A public contract reveals its inputs. Many agreements — a credit check, a medical rule, a sealed bid — cannot be published in order to be settled. Enigma's proposal is to split data across nodes, compute on the shares, and reveal only the agreed output.
What the design proposes
- Secret sharing splits an input so no single node holds it.
- A computation protocol produces an output without reconstructing the input on one machine.
- A public ledger can store the resulting commitment or payment while the data stays in the private layer.
How the mechanism is specified
- The privacy claim is as strong as the sharing threshold and the honesty assumption in the protocol.
- Performance and the class of programs that are practical are constraints the paper is working under, not details to skip.
- Secret Network's later trusted-execution design is a different technical bet. Do not cite Enigma as if it specified that implementation.
What this page does not treat as proven
- This is not the Secret Network documentation.
- Multi-party computation among a small, static set is a different trust model from a public chain.
- The paper does not make every smart contract private.
Why the desk still reads it
The right citation when a venture says 'private smart contracts' and means secret sharing, as opposed to a proof, a trusted enclave, or simply an access-control list. Those four are different products. Enigma is the first.
09 Lexicon
Terms, opened into the record
- Secret sharing
- Splitting a value into shares so a threshold can reconstruct it and a smaller set learns nothing.
- Multi-party computation
- A protocol in which parties compute a function of their private inputs without revealing those inputs.
- Threshold
- How many shares, or nodes, are required before the secret or the result can be produced.
- Coordinator chain
- The public ledger that records requests and settlement, not the private data.
10 Repository
Every linked record on this page
Underlined words open a page that already exists: a concept, a protocol profile, a failure record, or another paper. If a word is not underlined, this desk does not have a record for it.
This is Blockchain Lab's reading of a public design paper. It is not the paper, not a copy of it, and not an offer of tokens, equity, custody or a partnership. Later network behaviour can diverge from the text. Nothing here is investment, legal or technical advice.
Research status: Design paper. Last reviewed: 1 October 2026. This is a reading of a public paper, not investment, legal or security advice.