← Back to Master Map

Machina iO

Practical indistinguishability obfuscation. 2026 focus: noise refreshing, blind PRF over key-homomorphic encodings, ≥64-bit obfuscation, SNARK verification kickoff.

Research
Roadmap Completion: 10%
iO
GGH15
Lattice
FHE
key-homomorphic
Now

FHE multiplication over encodings

Implement FHE multiplication over key-homomorphic encodings. Foundation for blind PRF; unlocks predicate encryption / LFE.

Q1 2026 · In progress

Noise refreshing + dummy blind PRF

Implement noise refreshing of GGH15 encodings with replaceable dummy blind PRF. Confirm parameter growth is polylogarithmic.

Q2 2026

Benchmark harness

Circuit size/depth sensitivity and parameter-growth behavior. Set targets for real blind PRF circuit size.

Q2 2026

Noise refreshing paper

Paper: noise refreshing construction and security proof (venue TBD).

Q2 2026
Next

Blind PRF over key-homomorphic encodings

Circuit over encodings that simulates a PRF without revealing key or output. Replace dummy in noise refreshing.

Q2–Q3 2026 · Critical

≥64-bit obfuscation

End-to-end obfuscation and evaluation for ≥64 input bits. First practical iO beyond lookup-table scale.

Q3 2026

Devcon 2026

Paper and presentation: first practical-performance iO for nontrivial input size.

Q3 2026
Later

SNARK verification over encodings

Milestone 5 kickoff: PV vs DV scheme selection, verification circuit over key-homomorphic encodings. Continues into Q1 2027.

Q4 2026

Collaboration: security · efficiency

Academic collaboration on cryptanalysis (all-product LWE, evasive LWE, encodings) and efficiency improvements.

2026

Description

  • Execute 2026 critical path toward practical iO
  • Noise refreshing in practice; real blind PRF; ≥64-bit obfuscation
  • SNARK verification over encodings (kickoff Q4)

Deliverables

  • FHE multiplication + noise refreshing (dummy then real blind PRF)
  • First ≥64-bit obfuscation with reproducible benchmarks
  • Noise refreshing + security proof paper; Devcon 2026 dissemination
  • SNARK verification scheme selection and early prototype (Q4)

Impact

  • First practical-performance iO for nontrivial input size
  • Foundation for predicate encryption / LFE implementations
  • Security and efficiency collaboration with academia