Privacy in Crypto Is Getting an Upgrade

New privacy technologies are emerging: MPC, $FHE, TEEs, and zkTLS. Each tackles privacy from a different angle — and major projects are building with them now.

Here’s how these technologies work — and where they’re going👇

~~ Analysis by @davewardonline ~~

Multi-Party Computation (MPC)

MPC allows multiple parties to compute something together without revealing their inputs.

Say six friends want to find their average salary without revealing individual amounts. Each splits their salary into six random shares and sends one to each person. Everyone holds a share from each person but can’t reconstruct any full salary. They compute on the shares, not the actual numbers. The results combine into the correct average without exposing inputs.

MPC is useful when regulation or competition prevents data sharing, but joint analysis is beneficial. A common example: hospitals wanting to train AI models on private patient data without sharing it directly.

◆ Roadblocks for MPC — More participants mean more communication and computation, which slows things down. Blockchains can penalize bad actors but can’t eliminate the heavy resource costs.

Who's Using MPC and for What?

→ @FireblocksHQ — Splits private keys for institutional custody
→ @ArciumHQ — Uses MPC for private AI and sensitive computing
→ @renegade_fi — Dark pool for confidential onchain trading

Fully Homomorphic Encryption ($FHE)

$FHE lets you compute on encrypted data without ever decrypting it.

Normally, data is encrypted in transit but decrypted for processing. With $FHE, encryption stays in place throughout.

Imagine sending a locked safe with programmable gloves. You put private data inside, add instructions like “add these” or “sort this,” and send the safe and gloves to someone else. They perform the operations without opening the safe, then return it. You unlock it and see the result.

◆ Roadblocks for $FHE — $FHE is slow, 10–100x slower than other computation. Adding zk verification (zkFHE) ensures correctness but makes it even slower. $FHE hides data but doesn’t prove the computation was done right. zkFHE fixes that at the cost of speed.

Who's Using $FHE and for What?

→ @zama_fhe — $FHE tooling for encrypted smart contracts
→ @FhenixIO — Brings $FHE to practical apps
→ @Privasea_ai — Uses $FHE to train encrypted AI models
→ @octra — General-purpose chain with FHE-based compute and ML consensus

Trusted Execution Environments (TEEs)

TEEs are secure hardware zones that isolate data from the rest of the device, including OS and operators.

iPhones use TEEs for biometrics. Face or fingerprint data lives in secure chips. When you authenticate, a new scan is compared inside the TEE, which returns a simple yes/no — no raw data leaves the chip.

In crypto, TEEs are used for private contracts and block production. @unichain, Uniswap’s L2, uses TEEs to prevent exploitative MEV.

◆ Roadblocks for TEEs — TEEs rely on hardware vendors, making them centralized by crypto standards. They’re vulnerable to supply chain attacks or flaws, like the Intel exploit that compromised Secret Network’s TEEs.

Who's Using TEEs and for What?

→ @SpaceComputerIO — Uses orbital TEEs for tamper-proof satellite nodes
→ @OasisProtocol — L1 running private EVM contracts inside TEEs
→ @PhalaNetwork — Confidential cloud using TEEs from various providers

Zero-Knowledge Transport Security Layer (zkTLS)

zkTLS merges TLS (used in HTTPS) with zero-knowledge proofs to verify data while keeping it private.

TLS secures 95% of web traffic. zkTLS lets anyone prove facts from this data — like a bank balance — without revealing details.

Say you want an onchain loan and need to prove your bank balance. A zkTLS tool accesses your bank over HTTPS, reads the displayed balance, and creates a proof. The DeFi lender sees that your balance exists but not the number or history.

◆ Roadblocks for zkTLS — It only works with visible HTTPS data. It depends on TLS standards staying in use and requires oracle involvement, adding latency and trust.

Who's Using zkTLS and for What?

→ @zkp2p — Private on/off ramp
→ EarniFi — Lending to employees based on earned wages
→ @daisypayapp — Influencer payouts verified through zkTLS

Each PET offers different strengths and trade-offs. Complex apps may use multiple — MPC for coordination, $FHE for compute, TEEs for key storage. Many zkTLS tools incorporate other PETs internally.

Together, these tools expand crypto’s design space. But adoption will hinge on improving the user experience of privacy itself.

— h/t to @milianstx for his piece on, "WTF is MPC, $FHE, and TEE?" which served as a great starter