I recently spoke with a colleague working on an open-source secure element chip. While the goal is admirable — enabling customers to verify designs for security vulnerabilities and backdoors — I’m skeptical of the value proposition despite the hype and investor attention they’ve received.

I get why people are so excited by the idea. The tech industry becomes more interested in security practices with each major incident. People and businesses alike are increasingly focused on trust models. Companies hire auditors, run bug bounty programs, and sponsor open-source development to minimize black-box dependencies. Cryptocurrencies have even made retail customers security-conscious, scrutinizing their software more and even reaching for hardware wallets to move their secret keys offline. An open source secure element would give people a hardware tool to secure their keys, without requiring them to trust yet another black-box in their tech stack… right?

Secure Element ASICs Require Trust

As someone who’s spent years building hardware wallets, RFID card applets, and cryptographic libraries, I understand both the appeal and limitations of open-source secure elements. Unfortunately, no amount of open-source design files can address the fundamental issue: you cannot verify the integrity of the semiconductor itself. You must still trust that:

  • The RTL sent to fabrication matches the open-sourced version
  • The fab doesn’t introduce defects or vulnerabilities during production
  • Your chip is genuine and from official sources

If you’re seeking trustlessness, or if your trust model otherwise precludes trusting the chip vendor then these trust assumptions are likely also unacceptable to you.

So who benefits from open-source secure element ASICs? The value proposition is unclear. Secure elements are designed for customers who accept trust in the chip vendor—they need protection from remote and opportunistic physical threats, not from advanced threats or supply chain attacks. For these customers, whether the design is open-source is largely irrelevant since they cannot verify the physical chip matches the published design anyway.

FPGA-Based Secure Elements

Despite my criticisms, I am still an avid advocate for open-source design. The ability to inspect and review a technology, before I use it, helps us minimize trust and improve our online security.

This is only possible if I am able to inspect that the finished product matches the open-source design. For example, if I can verify in any of the following ways:

  • Compile and deploy code from source
  • Inspect deployed code against source (as with ROM)
  • Compare the traces of a PCB to the schematic
  • Assemble my own PCB/circuits at home

Unfortunately, ASICs cannot be fabricated or verified from home. You just have to trust what’s in the black-box.

FPGAs offer a solution. With secure element RTL and a compatible FPGA, you can compile and deploy the bitstream yourself, ensuring the design matches your expectations.

Caveats exist—many secure element features cannot be reasonably provided by FPGAs (such as anti-tamper circuitry or PUFs). However, core cryptographic operations and key storage—sufficient for most use cases—work well on FPGAs. The Precursor demonstrates this approach in practice, using an FPGA to provide secure key storage and cryptographic operations with user-verifiable firmware.

Of course, FPGAs aren’t a perfect solution—you still must trust the FPGA manufacturer’s silicon. However, the trust model differs significantly: you control the bitstream compilation and deployment, and can verify what’s loaded onto the device. While an FPGA could theoretically contain backdoors that analyze and modify your bitstream during configuration, the computational complexity makes this impractical. Parsing and backdooring arbitrary RTL in real-time would require resources orders of magnitude beyond what’s available in the die area around the gate array.

Conclusion

Trusted secure elements remain valuable tools, but for those of us focused on trust minimization, FPGA-based solutions offer a compelling alternative worth exploring.