Fortinet and Post Quantum Cryptography: How to be Prepared

The power of quantum computers changes the encryption landscape. Migrating to algorithms that are quantum resilient is a non-negligible task for companies to tackle. Fortinet can help, from giving crypto visibility across your footprint to integrating crypto agility into our firewalls. At Security Field Day 16, Chris Borales and Wei Ling Neo of Fortinet explained how the company supports post-quantum cryptography (PQC) migration.

Borales, who handles product marketing for Fortinet’s network detection portfolio, warned that attackers are already harvesting data now to decrypt it later. He cited estimates that classical cryptography faces a 15% chance of being broken by 2026, roughly 75% by 2030, and near certainty by 2034. Organizations that keep data for long periods, such as healthcare, finance, and government, face the greatest risk. FortiNDR Cloud builds a cryptographic inventory without decrypting traffic. It reads the handshake negotiations that happen in the clear and reports TLS and SSH PQC compliance, weak ciphers, and servers that ignore PQC-capable clients, with drill-down to individual IP addresses. Neo, who covers FortiOS and FortiGate in product management, pointed to US deadlines of the end of 2030 for quantum-safe key exchange and 2031 for digital signatures, with similar timelines in the EU and Australia. FortiManager tracks crypto inventory and configuration drift across the FortiGate fleet. Neo said drift often appears when security teams forbid fallback to classical algorithms but operations teams relax that setting to restore broken VPNs. FortiAnalyzer reports which TLS 1.3 sources and destinations are not using PQC. In IPsec testing, ML-KEM-768 offered the best balance of security and performance, while FrodoKEM and BIKE took more than three times as long to set up tunnels. ML-DSA certificates roughly halved performance compared with ECDSA. Fortinet’s NP7 ASIC accelerates traffic after the initial key exchange, and a demo showed more than 2,000 tunnels running with ML-DSA authentication.

Neo stressed that PQC carries real performance costs, so certificate optimizations are likely as the industry adapts. Fortinet has tested interoperability with strongSwan and will not fall back to classical algorithms when PQC is explicitly configured. Asked about a recently reported AI-discovered weakness in the Falcon algorithm, Neo said none has been found in ML-KEM. He added that defense should be layered, combining classical and quantum-safe algorithms and eventually adding quantum key distribution, which relies on physics rather than mathematical difficulty.