When NIST published ML-KEM, ML-DSA and SLH-DSA as Federal Information Processing Standards 203, 204 and 205 in August 2024, the debate over which algorithms would replace today’s public-key cryptography largely came to an end. For security leaders, the question has since shifted from whether to migrate to how quickly it can be done. Attention is increasingly turning to application programming interfaces (APIs), which now carry much of the sensitive data exchanged between businesses, cloud services and their customers.
Most APIs today terminate TLS connections using classical key exchange based on RSA or elliptic-curve cryptography. That traffic is considered secure against today’s computers, but it can be captured and stored now, then decrypted once a cryptographically relevant quantum computer becomes available. Security agencies refer to this as “harvest now, decrypt later.” The concern is greatest for data that must remain confidential for many years, such as health records, financial information, government communications and intellectual property.
Researchers often explain the risk with a simple test associated with Canadian quantum security researcher Michele Mosca. An organization is already exposed if the number of years its data must stay confidential, plus the number of years migration will take, is greater than the number of years until quantum computers can break current encryption. For sectors with long-lived data and slow procurement cycles, many analysts believe that line has already been crossed.
Regulatory timelines are adding pressure. The U.S. National Security Agency’s Commercial National Security Algorithm Suite 2.0 (CNSA 2.0) sets migration timelines for national security systems. Related obligations under frameworks such as FedRAMP, FISMA and HIPAA can extend to vendors and service providers throughout the supply chain. Industry observers note that long-lived data left exposed today may become an audit finding tomorrow.
Migration also raises practical concerns. Organizations relying only on classical algorithms remain exposed to future quantum attacks. Moving entirely to newly standardized post-quantum algorithms, however, means relying on schemes with a much shorter real-world track record. As a result, many security bodies and technology providers have adopted hybrid key exchange, which combines a classical algorithm with a post-quantum one, so a connection stays secure as long as either algorithm holds.
The software that carries encrypted traffic is another factor. Some of the most serious TLS vulnerabilities of the past decade, most notably Heartbleed in 2014, were memory-corruption bugs in software written in C. The NSA and CISA have since urged vendors to adopt memory-safe programming languages. Many organizations also lack a clear inventory of which cryptographic algorithms are used across their systems, which makes migration planning difficult. Changing a cipher has traditionally required significant engineering work rather than a configuration change.
Canadian quantum security company SuperQubit Inc. has developed SQ QuantumEdge, a post-quantum REST API platform designed to address these challenges. According to the company, the platform protects API sessions with ML-KEM key establishment and ML-DSA signatures, using a hybrid handshake that keeps a classical algorithm alongside the quantum-resistant one. Its cryptography runs on AWS-LC, Amazon’s open-source cryptographic library, whose FIPS 140-3 validated module includes ML-KEM. Because ML-KEM is now NIST-approved, the hybrid handshake can operate within FIPS-approved key establishment.
Both the load-balancing edge and the API server in SQ QuantumEdge are written in Rust, so the full request path is memory-safe. The edge is built on Pingora, the open-source Rust framework that Cloudflare uses to handle a very large share of its global traffic. The company says the platform also uses constant-time implementations to resist timing attacks and clears key material from memory after use. Supply-chain questions are addressed through a signed software bill of materials (SBOM), audited dependencies and reproducible builds.
To support crypto agility, SQ QuantumEdge applies policy-driven algorithm selection at both the edge and the server, so organizations can adopt or retire an algorithm through configuration rather than re-architecture. A cryptographic bill of materials (CBOM) shows which algorithms are in use across an organization’s API surface, providing continuous evidence for auditors instead of a once-a-year review. Because the handshake is hybrid, the company says existing clients continue to connect normally, and post-quantum protection can be enabled one endpoint at a time.
SuperQubit identifies government and public sector agencies, defense and aerospace, financial services and payments, healthcare and genomics, energy and critical infrastructure, and telecommunications as the sectors with the most immediate need. All of them handle data that must remain confidential for decades and operate under strict regulatory oversight.
With the standards now finalized and regulatory deadlines approaching, post-quantum migration is moving from long-term planning into operational reality. For organizations whose APIs carry long-lived sensitive data, the time available to close the “harvest now, decrypt later” window may be shorter than many expect.