Schedule

Quantum computing threatens to transform the foundations of modern cryptographic security, putting widely deployed algorithms such as RSA, Diffie-Hellman, and ECDH at risk while creating an immediate “store now, decrypt later” threat. This panel examines how organizations can prepare for the transition to post-quantum cryptography before Q-Day arrives. Experts from academia, government, and industry will discuss the technical, engineering, assurance, validation, and deployment challenges of migrating to quantum-resistant security across software, protocols, hardware, and cryptographic modules. Against the backdrop of Executive Order 14412, the panel will explore what organizations should be doing today to prepare for the post-quantum era.

Confidence in cryptography depends on correct implementation, reliable entropy, disciplined key management, and evidence of secure operation. This panel introduces the cryptographic validation ecosystem maintained by NIST and the Canadian Centre for Cyber Security, explaining how algorithm validation, entropy assessment, and FIPS 140-3 module validation work together. It examines emerging challenges, including complex supply chains, cloud deployments, faster product releases, and migration to post-quantum cryptography. Attendees will learn how validation serves as a repeatable assurance discipline that reduces implementation risk, supports interoperability, and strengthens trust in cryptographic products and services throughout their development, deployment, and ongoing use in operational environments.

Private information retrieval (PIR) enables clients to retrieve database records without revealing which records they request. It supports privacy-preserving applications including certificate transparency, password breach monitoring, media streaming, web search, and caller ID lookup. This talk surveys recent advances in practical, efficient PIR, beginning with an overview of state-of-the-art constructions. It then explores how hardware acceleration can improve scalability and concludes with emerging research on doubly-efficient PIR.

Cryptographic security ultimately depends on the unpredictability of cryptographic keys, and that unpredictability begins with a trustworthy noise source. This panel explores the physical and non-physical sources used to generate entropy for secure random bit generation, examining their strengths, limitations, and potential failure modes. Experts will discuss entropy modeling and estimation, statistical and health testing, conditioning, portability, and validation. The discussion will compare approaches under NIST SP 800-90B and SP 800-90C and BSI AIS 20/31, exploring where these frameworks converge and differ. Ultimately, the panel asks: What gives us confidence in the randomness behind our keys?

Many interactions between modules are driven by non-human actors. For example, a TPM is often used to unlock a boot disk at machine startup. Or an integrated root of trust is used to attest to a device’s integrity when starting a VM. ISO 19790 Level 2+’s authentication requirements do not support these use-cases without unnatural module boundary definitions that do not align to commercial offerings. As a result, Level 2+ is increasingly being viewed as not a viable target for a number of market segments. In this talk we will describe several aspects of this problem and possible solutions.

Quantum computers will eventually break the public-key cryptography protecting today’s systems. NIST has finalized three post-quantum standards — ML-KEM, ML-DSA, and SLH-DSA — but adopting them is proving harder than choosing them. While the public web experiments with shortcuts like Merkle Tree Certificates, government and regulated systems must deploy only algorithms validated under FIPS 140-3. This talk offers a practical roadmap: what the new standards require of vendors, the state of ACVP testing, the migration deadlines set by NIST and NSA, and the pitfalls that delay validations — and why post-quantum migration will amplify them — drawn from CMVP laboratory experience.

In 2022, attackers exfiltrated over 25 million LastPass vaults. No algorithm had failed. AES-256 performed exactly as specified. This presentation diagnoses three failures against FIPS 140-3 requirements: key derivation left at 5,000 PBKDF2 iterations on legacy accounts, a cryptographic boundary never drawn around plaintext URLs, and backup decryption keys held without enumerated access rights. Each violated a property that module assurance makes testable and commercial practice leaves implicit. The session considers whether cryptographic assurance should condition market access for products holding sensitive data at scale.

The transition to post-quantum cryptography raises an urgent question: can cryptographic testing and validation keep pace? This talk examines bottlenecks in the FIPS 140-3 validation pipeline and efforts by NIST, NCCoE, and industry practitioners to replace labor-intensive processes with automated workflows. It explores protocols for exchanging evidence supporting entropy source, algorithm, and module validation, while highlighting the need to improve workflows across all stakeholders. Focusing on Intelligent Automation (IA), the presentation outlines a path toward faster, repeatable, computer-driven validations and considers the work still needed to prepare the validation ecosystem for the demands of widespread post-quantum adoption across products and systems.

Artificial intelligence is rapidly reshaping cybersecurity, creating powerful new capabilities for both defense and attack. Frontier AI models can increasingly discover vulnerabilities, develop exploits, conduct complex cyber operations, and even contribute to cryptanalysis. As these systems become more autonomous and capable, they also introduce new risks to software, infrastructure, and cryptographic security. This panel brings together experts from academia, government, and industry to explore the evolving intersection of AI and cybersecurity. The discussion will examine AI-enabled cyber operations, cryptographic vulnerabilities, agentic AI, governance, product security, and enterprise risk, asking a critical question: How will AI change the security of our digital world?

ISO/IEC 19790:2025 arrived in 2025. Ryan Thomas sat on both sides of the validation table, first as a laboratory director and now as a vendor guiding NVIDIA products through FIPS 140-3. This talk presents a top 10 list for selective adoption: adopt the improvement, not the edition, and a path that captures 2025 updates through CMVP’s own instruments (Implementation Guidance, SP 800-140, templates, and profiles) while FIPS 140-3 stays anchored to the 2012 edition.

The transition to post-quantum cryptography requires more than adding PQC algorithms to HSM APIs. HSMs both provide and depend upon cryptography, so quantum resilience must extend throughout the architecture, including roots of trust, secure booting and updating, attestation, authentication, communications, auditing, and backups. This talk examines architectural approaches for building quantum safe HSMs that can securely evolve as cryptographic algorithms and threats change using crypto agility, while eliminating vendor lock-in using crypto portability to securely migrate key material between platforms. The presentation also identifies gaps in FIPS 140-3 and proposes directions for addressing them.

Nearly every physical entropy source submitted for validation under SP 800-90B belongs to a small family of microelectronic noise mechanisms, creating a systemic risk due to a monoculture. This talk reports practical experience validating a macro-scale electromechanical entropy source. We cover the
fluid dynamics behind the noise source, the imaging and resulting raw noise samples, and the properties that matter for validation rather than for marketing.
We ask what changes when a centralized entropy source’s output is delivered to consuming endpoints as a metered service. Finally, most real deployments consume entropy through a cryptographic library rather than directly. We discuss integration into widely used open-source stacks, whose seeding paths and their entropy assumptions are documented in code.

Metastability-based digital noise sources are foundational IP used for high-performance hardware random number generation. They have higher output bit rates, simpler entropy justification, and resistance to published hardware attacks compared to ring oscillator noise sources. Recent implementations using contention-based CMOS bistable circuits, switched capacitor charge pumps, and novel self-clocking schemes have further advanced the state of the art. The well-known Markov Chain entropy justification for metastability-based noise sources also requires advancement as flicker noise becomes more prominent in advanced process nodes. Modeling the noise source as a delta modulator ADC is useful for understanding the design and the statistical properties of its output.

The central insight is that the hard work of FIPS 140-3 conformance for a complex platform is architectural and documentary, not cryptographic, and the anticipated takeaway is a concrete, experience-grounded framework for de-risking FIPS 140-3 transformation. Attendees will understand why financial HSMs have evaluation challenges that PKCS #11-based HSMs do not and how service indicators, SSP lifecycle rules, and mode transitions change. The presentation is structured as a four-part case study:

1. motivation and problem framing;
2. technical deep-dives on the three transformation areas;
3. lessons learned with specific examples; and
4. practical recommendations.

Caliptra’s open-source, multi-vendor development model challenges conventional FIPS 140-3 validation practices built around manual, single-vendor assessments. This talk compares the gap analysis and SP/MIS drafting process used for Caliptra V1.x with the structured, tool-supported, traceable process now applied to V2.x. Drawing on direct project experience, it examines parallel validation and analysis efforts, cross-organization review coordination, and the pursuit of full assessment coverage through publicly available implementation walkthroughs. Attendees will gain practical insights into improving repeatability, evidence reuse, and reviewer alignment for evolving open-source cryptographic modules.

Storing cryptographic identities in non-volatile memory (NVM) leaves hardware vulnerable to physical probing and memory extraction. Physically Unclonable Functions (PUFs) mitigate this threat by deriving unique device fingerprints on demand from inherent silicon manufacturing variations. However, NIST strictly requires primary cryptographic keys to originate from an approved Random Bit Generator (RBG), preventing standalone PUF responses from serving as primary FIPS-compliant keys.

This presentation outlines a compliant hybrid architecture that embeds PUF keys into security boundaries strictly as secondary parameters. We present concrete integration pathways across NIST SP 800-133, 800-132, 800-108, 800-38A, and CC/AIS 31 bijective mappings. Finally, we detail essential hardware validation methodologies, focusing on pre-silicon stochastic modeling, Error Correction Code (ECC) helper data entropy deductions, and silicon lifecycle testing.

Cryptographic standardization, certification, product lifecycle chains, and government mandates all operate on different timelines, and the multi-year engineering, validation, certification, and procurement process is already well underway. For products with years’ long development and operational lifecycles, waiting for greater certainty about the quantum threat may mean waiting too long to meet the deadlines that are already certain.

Email remains essential to modern communication, yet achieving FIPS 140 compliance for the widely used OpenPGP and S/MIME protection protocols presents significant challenges. This presentation examines their required and optional cryptographic algorithms, identifies gaps between current standards and FIPS 140 requirements, and explores practical approaches for addressing them. Using GnuPG and its cryptographic library libgcrypt as real-world implementations, the presentation will demonstrate configuration strategies for FIPS-compliant OpenPGP and S/MIME while considering interoperability. The presentation will also examine the transition to post-quantum cryptography, emerging standards, and how PQC algorithms may be incorporated into future FIPS-compliant email protection.