A brief history of the U.S. trying to add backdoors into encrypted data (2016)

Original Article ↗ Hacker News Discussion ↗

US efforts to weaken or bypass encryption—from Cold War-era manipulation of commercial ciphers and export controls on “strong” cryptography to modern pushes for lawful-access backdoors—are examined through both historical examples and recent revelations. Commenters highlight how hardware features (Intel ME, AMD PSP, secure enclaves), opaque firmware, and government–vendor partnerships can create hidden vulnerabilities, while stressing that any master key or backdoor inevitably becomes a high‑value target for criminals and foreign states. The thread also touches on current pressure to undermine end‑to‑end encryption in the name of fighting terrorism and child abuse, and on the difficulty of meaningfully verifying that chips, phones, and services are free of covert access paths.

Hardware / Firmware Backdoors and Blobs

  • Intel ME and AMD PSP are seen as large opaque blobs running below the OS.
  • Some argue they’re a “gaping hole”; others say they’re just one of many risky firmware blobs (NIC, Wi‑Fi, GPU, EFI) and often not the worst.
  • There are claimed ways to partially/mostly disable ME/PSP (MEcleaner, HAP bit, disable commands), and many consumer systems lack the full vPro/remote‑management wiring.
  • Concern extends to all embedded firmware: remotely exploitable Wi‑Fi, GPU, UEFI bugs, and EFI implants used for persistence.

Trust in Hardware Vendors (Intel, AMD, Apple, RISC‑V)

  • Consensus: you ultimately must trust your CPU/SoC vendor; if they want a backdoor, they can hide it in silicon.
  • Apple Silicon is viewed by some as relatively hard to backdoor completely, but a recent cache/MMIO “test register” bug shows powerful hardware flaws can exist for years and be abused remotely (e.g., via iMessage).
  • RISC‑V is welcomed as an open ISA, but many concrete implementations and GPUs/AI accelerators remain proprietary; open design CPUs do exist but are still maturing.

Government Master Keys and Crypto Backdoors

  • Proposal: a “sealed box” design with a government‑held master private key could, in theory, be safe if the key never leaks.
  • Strong pushback: master keys and “golden keys” leak repeatedly (examples: Microsoft signing keys, Realtek, DVD/Blu‑ray, HDMI, TSA locks, crypto exchanges).
  • Critics emphasize: any such key must stay secure for decades, be accessed widely, and would be the single most valuable target; eventual compromise is seen as inevitable.
  • Others note added implementation risk: test keys, misconfigurations, and protocol bugs around the escrow mechanism itself.

Historical and Structural Backdooring

  • Crypto AG: long‑running CIA/BND ownership and weakened products used by many governments.
  • NSA influence on algorithms and hardware: DES key‑size and export limits; “SIGINT‑enabled” encryption chips; weakened or co‑designed commercial VPNs and OS components; various documented backdoored systems.
  • Crypto was historically treated as a munition (ITAR), enabling businesses like Thawte; PGP’s printed‑book export and “munitions t‑shirts” are cited as protests that helped establish “code as speech”.

Law‑Enforcement Narratives and CSAM

  • Article context: post–San Bernardino, FBI pushing for iPhone access.
  • Commenters note repeated use of terrorism and later CSAM as arguments for encryption backdoors and on‑device scanning.
  • Many see this as a powerful political wedge: opposition risks being framed as pro‑criminal, despite backdoors’ systemic risk.

Threat Models and Practical Security

  • Several argue governments can already get most targeted data via device exploits, supply‑chain attacks, “wrench” tactics, or legal compulsion, without universal backdoors.
  • Many prioritize protection from corporations, cybercriminals, and mass surveillance over trying to be government‑proof, which is often viewed as unrealistic with modern hardware side channels.