Cloudflare maps hidden cryptography with an internal AI tool

Cloudflare is using an internal AI-assisted tool called CryptoLabe to find where cryptography appears across its codebase and identify dependencies that complicate upgrades. The inventory is part of the company’s target to make its platform fully ready for post-quantum cryptography by 2029.

Many Cloudflare products already use post-quantum encryption with TLS 1.3, but remaining work includes older TLS connections, other uses of public-key encryption and post-quantum authentication. CryptoLabe is intended to give engineering teams repository-level guidance, count classical and post-quantum uses, and surface protocols or libraries that must change first.

Large migrations often fail at discovery: code can call cryptographic libraries indirectly or depend on standards that are not ready. AI can help classify that long tail, but findings still require validation before engineers alter security-sensitive code. Cloudflare treats the model as a mapping assistant, while progress metrics and technical review remain central.