How Quantum Snake Oil Vendors Respond When You Ask Hard Questions

This article is part of the Quantum Snake Oil Dictionary — a series examining terms used in quantum technology marketing. This is the companion guide: after you’ve asked the hard questions, here’s how to read the answers.

Introduction

Throughout the Quantum Snake Oil Dictionary, I suggest specific questions to ask vendors who use suspicious terminology. Questions like: What algorithm does your product implement? Has it been submitted to NIST’s standardization process? Can you point me to a peer-reviewed paper describing the security proof?

These are straightforward technical questions. Any company building genuine quantum or post-quantum security technology will find them routine, even welcome. But if the vendor is peddling something questionable, your questions will trigger a very different kind of response.

Over the years, I have asked these questions many times, and I have heard from readers and colleagues who have asked them on my behalf. What follows is a field guide to the response patterns I have observed. It is not a comprehensive taxonomy of every deflection technique, but it covers the most common ones.

The core insight is simple: a legitimate company responds to a technical question with a technical answer. A questionable company responds to a technical question with anything other than a technical answer.

The Deflection Playbook

1. “You’re Funded by Big Tech to Suppress Innovation”

This is the most common deflection, and it comes in several flavors. The claim is that anyone asking critical questions is secretly working on behalf of IBM, Google, the World Economic Forum, or some other large entity that supposedly wants to protect the status quo.

I can speak to this one personally. My background includes work at IBM, PwC, KPMG, and other large organizations, and I make no secret of it. This makes me an easy target for the accusation. The logic goes: I must be attacking independent innovators because I am a puppet of the establishment.

The tell is that the accusation replaces the technical answer. If someone asks “does your system produce Bell-violating correlations?” and the response is “you’re asking that because IBM pays you to suppress disruption,” the response has revealed that there is no technical answer available.

A legitimate company does not care about my CV. They care about the technical question, and they answer it.

2. “The Scientific Community Suppresses Disruptive Ideas”

This is the Galileo gambit — the claim that the entire physics and cryptography community is a closed cartel that rejects genuinely revolutionary work to protect its grant funding and institutional prestige.

It is true that the scientific community can be slow to accept new ideas, and there are historical examples of initially rejected work that proved correct. But there is a significant difference between a researcher who submits work to peer review, receives critical feedback, and iterates — and a vendor who has never submitted their claims to any form of independent scrutiny and accuses the entire field of corruption when asked why.

NIST’s post-quantum cryptography standardization process is specifically designed to evaluate new cryptographic approaches. It is open to anyone. Submissions are analyzed by the global cryptographic community over multiple rounds spanning years. The process that selected ML-KEM (formerly CRYSTALS-Kyber), ML-DSA, and SLH-DSA began in 2016 and involved contributions from researchers at universities, startups, and yes, large companies. If a vendor has a genuinely novel quantum-safe algorithm, the path to credibility runs through public cryptanalysis. If they refuse to take it, ask why.

3. “We Can’t Share Technical Details Because of Patents / NDAs / National Security”

This sounds reasonable on the surface. Companies do protect intellectual property, and some security work does involve classified applications. But there is a critical distinction between protecting implementation details and refusing to name the underlying cryptographic approach.

A legitimate vendor will tell you: “We implement ML-KEM-1024 for key encapsulation, with a hybrid X25519 classical fallback, compliant with NIST FIPS 203.” The implementation details (optimized code paths, hardware acceleration, key management architecture) might be proprietary. The choice of algorithm never should be.

This is not my personal opinion. It is Kerckhoffs’s principle, formulated in 1883: a cryptographic system should be secure even if everything about the system, except the key, is public knowledge. Any vendor who claims their algorithm itself must be secret for security reasons is violating a foundational principle of cryptography that has held for over 140 years.

Classified government applications do exist, but a vendor pitching to commercial CISOs while claiming they cannot discuss their cryptographic approach because of national security is almost certainly deflecting, not protecting state secrets.

4. “You Just Don’t Understand the Physics”

This is the intellectual intimidation play. Instead of explaining how their technology works, the vendor implies that your question reveals ignorance. If only you understood quantum mechanics at a sufficiently deep level, you would see why their product works.

This is particularly effective when the questioner is a security professional rather than a physicist, because it exploits the knowledge gap between cybersecurity and quantum physics. It is designed to make you feel unqualified to evaluate the claim.

The antidote: in legitimate quantum technology, the ability to explain how a product works to a technically literate non-specialist is considered a basic professional competence. If I cannot explain to a CISO how QKD provides security (photons, measurement disturbance, statistical testing for eavesdropping), then I have failed at communication, not demonstrated superiority. The physics of quantum security is subtle, but the product claims should be explainable. If a vendor cannot explain their product to you, the most likely reason is not that you are too ignorant. It is that the product does not do what they claim.

5. “Our Patents Prove the Technology Works”

A patent proves that someone filed paperwork with a patent office. It does not prove that the described invention works, that it is secure, or that it does what the marketing claims. Patent examiners evaluate novelty and non-obviousness, not physical correctness or cryptographic security.

The patent offices of the world contain filings for perpetual motion machines, faster-than-light communication devices, and methods of transmuting lead into gold. A patent for “quantum binary hybridization” or “simulated entanglement-based encryption” tells you that the application met the formal requirements for filing. It tells you nothing about whether the described system provides any security whatsoever.

A legitimate vendor may hold patents. They will also point you to peer-reviewed publications, independent security audits, or FIPS 140-3 validation certificates. The patents are intellectual property protection, not evidence of efficacy.

6. “Our Clients Include Governments and Defense Ministries”

Name them. Or at minimum, describe the use case with enough specificity that the claim can be investigated. “We work with defense clients” is unfalsifiable and costs nothing to say. “Our product is deployed in the Spanish Ministry of Defense’s classified communications network under contract reference X” is a verifiable statement.

Legitimate vendors with government clients will typically name them in case studies (with permission), or at minimum have publicly announced contract awards, press releases from the government agency, or references in procurement databases. Some classified work genuinely cannot be discussed, but a vendor whose entire evidence base is a vague claim of unnamed government customers is waving a red flag.

7. “We Won [Unverifiable Award]” and the Pay-to-Play Credibility Machine

Legitimate awards in quantum computing and quantum security come from recognized institutions: the European Physical Society, the Institute of Physics, the American Physical Society, IEEE, XPRIZE, and similar organizations. These awards have public records, named selection committees, and documented criteria.

If a vendor claims an award you cannot verify through the awarding body’s public records, that is a problem. If the award comes from an organization that appears to exist solely to give awards to the vendor, that is a bigger problem.

But the more subtle version of this is the pay-to-play ecosystem. There is a cottage industry of cybersecurity and technology magazines that will put anyone on their front cover, name them “CISO of the Year,” “Quantum Innovator to Watch,” or “Top 10 Cybersecurity Visionary” for a fee of $2,000–$5,000. The editorial process involves receiving payment and publishing the profile. These publications have professional-looking websites, PDF magazines with glossy layouts, and LinkedIn-friendly cover images that look impressive when shared.

The test is straightforward: is the publication one that practitioners in the field actually read and respect? If it is not a name recognized in the security community (IEEE, ACM, NIST publications, established trade press like Dark Reading, SC Magazine, or Wired), treat the “award” as a marketing expense, not a credential. A “Top 50 Quantum Leaders” list from a publication you have never heard of tells you that someone wrote a check, not that someone built a product.

Trade show awards operate similarly. “Best Product at [Conference]” sounds impressive, but many trade shows offer “best of show” or “innovation award” programs where exhibitors submit applications, sometimes with a fee. The evaluation is often cursory, and in some cases the award goes to every exhibitor who applies. Winning “Best Product” at a niche defense expo does not constitute independent technical validation.

A legitimate company may collect these accolades too. The difference is that they do not use them instead of technical evidence. If the press clippings and award plaques are the headline, and the peer-reviewed papers and FIPS validation certificates are absent, the credibility is manufactured.

8. “Our Research Is Published in Peer-Reviewed Journals”

This one requires particular care, because “peer-reviewed journal” sounds like the gold standard of scientific credibility. And for journals like Nature, Science, Physical Review Letters, IEEE Transactions, or the established cryptography venues (EUROCRYPT, CRYPTO, ASIACRYPT, Journal of Cryptology), it is.

But there is a parallel universe of predatory and pseudo-scientific journals that exist to provide the appearance of peer review without the substance. Some are straightforward pay-to-publish operations: submit a paper, pay a fee, receive a DOI and a citation you can put on your website. Others are more insidious — set up by a small group specifically to validate their own claims. The editorial board consists of the authors’ colleagues, the reviewers are sympathetic, and the “peer review” is a formality. The result is a closed credibility loop: the group publishes in its own journal, cites those papers as evidence, and presents the citations as independent validation.

If a vendor points you to a published paper, check three things. First, the journal: is it indexed in established databases (Web of Science, Scopus)? Is it listed on Beall’s List of predatory journals (or its successor, Cabells Predatory Reports)? Would a researcher in the field recognize its name? The Think. Check. Submit. checklist is a practical tool for evaluating unfamiliar journals. Second, the authors: do they have affiliations with recognized universities or research institutions, and do they publish in other venues as well? Third, the citations: has anyone outside the authors’ immediate circle cited the work?

Pre-print servers add another layer of complexity. arXiv is the standard pre-print server for physics, mathematics, and computer science. It has an endorsement system — new submitters need an endorsement from an established arXiv author in the relevant category — and papers posted to arXiv are discussed and scored by the research community on SciRate. An arXiv paper is not peer-reviewed, but it has passed a basic filter and is exposed to community scrutiny.

Other pre-print servers, such as preprints.org, have no such checks. Anyone can post anything. A paper that appears only on a no-filter pre-print server, and has not been posted to arXiv or submitted to a recognized journal, should be treated with appropriate skepticism. Ask the vendor: why is this paper not on arXiv? If the answer involves complaints about arXiv’s gatekeeping, that tells you something.

9. “We Have a Strategic Partnership with [Major Company]”

Using a major company’s product does not make you their strategic partner. Running your application on AWS does not mean you have a partnership with Amazon. Deploying your software on phones managed by BlackBerry UEM does not mean BlackBerry has validated your quantum technology. Listing your product on a cloud marketplace does not constitute an endorsement by the marketplace operator.

Some large technology companies publish customer case studies or success stories as part of their marketing programs. These are typically written based on information provided by the customer, and they describe how the customer used the large company’s platform, not an independent assessment of the customer’s product claims. A case study titled “How [Startup] Used [Big Company’s] Cloud to Deploy Their Quantum Security Solution” tells you that the startup is a customer of the cloud provider. It does not tell you that the cloud provider has validated the startup’s quantum security claims.

The test: does the “partnership” announcement come from both parties? Is there a joint press release from the major company? Does the major company’s own website describe the relationship? If only the startup is calling it a partnership, it may be a customer relationship being presented as a collaboration between equals.

10. Inflated or Unverifiable Credentials

“Dr.” is a title that carries weight, and rightly so. But not all doctorates are equivalent. A PhD in quantum information science from MIT, Oxford, or ETH Zurich, with a dissertation committee of recognized researchers and published papers in established journals, is a strong credential. A doctorate from an unaccredited institution, or a degree in an unrelated field (computer systems engineering, management) presented as though it confers expertise in quantum physics, is a different thing entirely.

Similarly, academic titles like “professor” mean different things in different contexts. A tenured professor at a research university holds a position earned through sustained peer-reviewed research. A visiting lecturer at a private institution may hold the same title through a less rigorous process. The title alone does not tell you which you are looking at.

I do not raise this to be dismissive of people from non-traditional academic backgrounds. Some of the best work in security and cryptography comes from practitioners without formal physics training. But if a vendor’s primary credibility claim is an academic credential, verify it: is the institution accredited? Is the degree in a relevant field? Does the person have a publication record in recognized venues? If the credential is the centerpiece of the pitch and it cannot be verified, treat the rest of the claims accordingly.

11. “We’re Backed by [Major VC Firm]” or “We Raised $X Million”

Venture capital funding from a recognized firm does carry some signal. Tier-one VCs like Andreessen Horowitz, Sequoia, or In-Q-Tel conduct due diligence before investing. A Series B or C round from a major firm means that professional investors, often with access to technical advisors, decided to put significant capital at risk.

But the signal is weaker than most people assume, and it degrades at earlier stages. A pre-seed or seed check of $500K–$2M involves far less scrutiny than a $50M growth round. At the earliest stages, VCs are betting on teams, markets, and narratives, not on validated technology. They accept that most of their portfolio companies will fail. A seed investment is a bet that the idea is worth exploring, not a certification that the technology works.

Even at later stages, VC due diligence evaluates the business case (market size, competitive positioning, team quality, revenue traction), not the cryptographic security of the product. A VC firm is unlikely to have commissioned an independent cryptanalysis of the vendor’s algorithm before investing. And VCs do make mistakes. The history of technology investing is littered with heavily funded companies that turned out to have fundamental technical problems.

The test: funding validates that professional investors believe the company has commercial potential. It does not validate that the product provides the security properties it claims. These are different questions, and they require different evidence.

12. “We’re Part of [Prestigious Accelerator / Innovation Center / University Program]”

Accelerators and innovation centers (Y Combinator, Techstars, university-affiliated incubators, regional innovation hubs) provide valuable support to startups. But the bar for admission varies enormously. Y Combinator is highly selective; a regional innovation center may accept any company that applies and pays a membership fee.

The credibility-laundering move is to present accelerator membership as institutional endorsement. A company that joined an innovation center’s accelerator program and uses its co-working space will market itself as “associated with” or “supported by” the institution, implying that the institution has validated their technology. In most cases, the institution has provided workspace, mentorship, and networking, not a technical assessment of the company’s product claims.

University associations work similarly. “Spun out of [University]” can mean that a tenured professor’s lab spent years developing the technology and the university’s technology transfer office licensed it. Or it can mean that someone who once attended the university is now using the affiliation in their marketing. These are very different things.

The test: is the relationship an active technology license, a co-development agreement, or a formal research collaboration with named faculty? Or is it an accelerator membership, a co-working arrangement, or a historical educational connection? The vendor should be able to describe the nature of the relationship precisely. If the description is vague, the relationship is probably thin.

13. “We’ll Demonstrate It to You Under NDA”

A private demonstration under NDA is not verification. A system that encrypts and decrypts successfully in a demo room proves that something is happening, but not that it provides quantum-grade security, or any specific security property. Any classical encryption system will also encrypt and decrypt successfully in a demo.

What a demonstration cannot show you is resistance to a specific attack model, correctness of the underlying mathematics, or absence of exploitable weaknesses. Those require formal security proofs and independent cryptanalysis, not live demos.

A legitimate vendor will welcome you bringing your own cryptographer to evaluate their claims. A questionable vendor will insist that only their own team can properly demonstrate the product.

14. Conference Speaking as Proof of Credibility

“Our CTO presented at [Quantum Conference].” This can mean many things. At established academic conferences (QIP, APS March Meeting, IEEE Quantum Week, PQCrypto), presentations are selected through competitive peer review of submitted abstracts or papers. Getting a talk accepted at QIP is a genuine credential.

At many industry and trade conferences, speaking slots are available for purchase. Sponsor packages routinely include a keynote or panel slot as part of the deal. Some events explicitly sell speaking opportunities as a product. Presenting at these events proves that someone paid a registration or sponsorship fee, not that their work was evaluated and accepted.

The distinction matters because “conference speaker” appears on LinkedIn profiles and company websites without any indication of whether the talk was peer-selected or purchased. If a vendor cites a speaking engagement as evidence of credibility, check whether the conference has a public program committee, a documented review process, and a track record of rejecting submissions.

15. Testimonials Instead of Specifications

“Our CEO was voted Top 50 Quantum Innovators.” “Our customers report 99.7% satisfaction.” “Our CTO was a keynote speaker at five quantum conferences this year.”

None of these address the question of what algorithm the product implements and what security properties it provides. Testimonials, magazine features, conference appearances, and customer satisfaction surveys are marketing activities. They are not evidence of cryptographic security.

A legitimate vendor will have testimonials and technical specifications. If the specifications are absent and the social proof is doing all the work, treat the claims with skepticism.

16. Moving the Goalposts

You ask about the cryptographic algorithm. They talk about energy efficiency. You ask about the security proof. They talk about ease of deployment. You ask about NIST standardization. They talk about their 25 years of experience.

Each answer individually might be reasonable in a different conversation. The pattern is the problem. When every technical question receives a response about a different topic entirely, the vendor is performing an evasion that is designed to exhaust your patience until you stop asking.

A simple test: after 15 minutes of conversation, have you received a direct answer to your original question? If not, you have your answer.

How Legitimate Companies Respond

For contrast, here is what I hear from companies that are building real quantum and post-quantum security products. This is not an endorsement of any specific vendor. It is a description of communication patterns that indicate technical substance.

They name their algorithms. “We implement ML-KEM-1024 for key encapsulation and ML-DSA-65 for digital signatures, per NIST FIPS 203 and 204.” No evasion, no jargon soup.

They acknowledge limitations. “Our QKD system requires a dedicated fiber link and currently operates at distances up to 100 km without trusted relays. For longer distances, we use trusted nodes, which introduces classical security assumptions at each relay point.” Honest limitation statements are a mark of technical seriousness.

They point to public evidence. “Our algorithm was submitted to NIST in Round 3 and was selected for standardization. Here is the submission paper. Here is the FIPS 140-3 module validation certificate number.” Evidence you can independently verify.

They welcome technical scrutiny. “Bring your cryptographer. We will walk them through our security proof and threat model.” Companies that are confident in their technology are not threatened by experts looking under the hood.

They distinguish marketing from engineering. Some legitimate companies do use simplified language in marketing materials. The difference is that when you scratch the surface, there is substance underneath. A glossy brochure that says “quantum-safe security” is fine if, when asked, the team responds with “specifically, we implement a hybrid X25519 + ML-KEM-768 handshake in TLS 1.3, with ML-DSA-65 for certificate signatures.”

They respond to technical questions with technical answers. This is the master test. Everything else flows from it.

The One Question That Reveals Everything

If you only have time for one question, ask this:

“Which specific NIST-standardized algorithm does your product implement, and do you have a FIPS 140-3 validated module?”

A legitimate vendor will answer in one sentence. A questionable vendor will answer with everything except the answer.