More Open Source Security episodes

Trusting your computing with Lance and Anton thumbnail

Trusting your computing with Lance and Anton

Published 9 Nov 2026

Duration: 38:21

Verifiable computing uses deterministic builds and hardware attestation to prove deployed software matches source code, mitigating supply chain attacks.

Episode Description

Josh chats with Lance and Anton from Caution about verifiable computing. We start out discussing all the challenges in bootstrapping software that alm...

Overview

The founders of Caution explain verifiable computing as a combination of reproducible builds and hardware-backed cryptographic attestation. They describe how deterministic compilation, starting from a small seed of machine code, allows verification that shipped binaries match their source code, addressing supply chain attacks. Their Linux distro StageX builds the entire operating system from zero with identical output every time, so that "we can be confident that this source code that we're looking at is what is in the output binaries."

Confidential compute, available in modern AMD, Intel, and Nvidia chipsets, provides memory encryption and cryptographic remote attestation. Combined with reproducible builds, this enables real-time proof of exactly what code runs on a server. The founders argue this can eliminate entire classes of attacks and that making deployment as simple as Docker containers is essential for widespread adoption, positioning the technology as "what comes after TLS."

What If

  • What if you could prove to your customers that your service doesn't log their data?

    • Move: Deploy your service (e.g., a VPN or proxy) in a confidential compute enclave and use cryptographic remote attestation to let customers verify the exact binary running on your infrastructure.
    • Why Now?: The founders identify this as a high-value use case: "I want to prove to my customers, maybe I'm a VPN provider and I want to prove to them that I'm not logging," and the tooling is being released as fully open source.
    • Expected Upside: As a possibility, customers gain verifiable trust without reading privacy policies, potentially differentiating your service in privacy-sensitive markets.
  • What if you could eliminate pre-generated bundle risk from your build pipeline?

    • Move: Restructure your CI/CD to pull source directly from version control and run all code generation (autotools, configure scripts) in a hermetic environment with a trusted toolchain, rather than consuming pre-built release tarballs.
    • Why Now?: The XZ incident demonstrated that "the source code that was actually reviewed in the repo was fine" but the pre-generated bundle was compromised, and the founders argue this is a widespread industry habit.
    • Expected Upside: As a proposal, this removes an attack surface where "somebody wants to hide a vulnerability, they will have to do it in the public source code that everybody gets to see."
  • What if you could sandbox AI agents with provable memory isolation?

    • Move: Run agent workloads in confidential compute enclaves with memory encryption, using attestation to verify the agent's code and boundaries before and during execution.
    • Why Now?: The founders explicitly mention "You can use this to sandbox agents more effectively" and "provably private AI inference" as use cases, and the technology is described as deployable via OCI containers without rewriting applications.
    • Expected Upside: As a possibility, you could offer provably private AI inference where users can verify "I can use LLMs and not pipe all of my personal data to a third party in a provable way."

Takeaway

  • Build from reviewed source in version control, not pre-generated release bundles; run all code generation steps yourself in a hermetic environment with a trusted toolchain.
  • Adopt a deterministic Linux distro like StageX as a drop-in replacement for your current base image to get reproducible builds without changing your development workflow.
  • Deploy containerized workloads in confidential compute enclaves using OCI-native containers; the founders confirm existing services "you could just run them in an enclave today" without modification.
  • Use cryptographic remote attestation to verify that deployed binaries match your source code, giving you and your users cryptographic proof of what is running.
  • Change one line in your Dockerfile to apply verifiable compute benefits to existing containerized applications, as the founders describe the deployment model as "you just change one line at the top of your file."

Recent Episodes of Open Source Security

5 Oct 2026 Dependency attacks in 2026 with James Matchett

"Software supply chain security faces evolving risks like install-time attacks, transitive dependencies, and AI-generated threats, requiring proactive monitoring, dependency firewalls, and regulatory compliance to balance security with developer flexibility."

21 Sept 2026 The curl summer of Bliss with Daniel and Stefan

"Explored open-source security challenges, highlighting curl's 'Summer of Bliss' initiative to reduce maintainer burnout, the impact of AI-generated vulnerability reports, and the need for human expertise in security management, while advocating for sustainable practices in open-source development."

More Open Source Security episodes