Can I Trust My Vendor’s Security Claims? Peer-reviewed vs. self-certification methods
Format-preserving encryption (FPE) is in the news recently, as two researchers demonstrated a cryptanalytic attack on one method that NIST had endorsed—FF3. NIST now expects to revise their endorsement of FF3 (Special Publication 800-38G) after details of the attack are published to either change the FF3 specification or withdraw approval. It’s important to be aware, this news is independent of NIST continued endorsement of FF1 format-preserving encryption (FPE).
However, this very review process—of a publication leading to expert analysis and subsequent revision for any newly discovered weakness—is precisely how we obtain trustable security systems. Without it, we must simply rely on obfuscation and hope from the words of sales reps, none of which is reliable for meeting the security requirements of today’s increasingly high-risk, technically-sophisticated world.
Nonetheless, since the FF3 attack was revealed a little over a month ago, fear, uncertainty, and doubt has started to emerge. Some of this is natural , as enterprises review options and understand impact. However opportunistic vendors might be attempting to re-direct the conversation to often-confusing alternatives to FPE. Perhaps worse, some may be denigrating the process of public review.
Proven in use by the largest organizations, FPE is an industry-defining breakthrough invention by HPE that has been securing the world’s most critical data, from financial information and health care data to sensitive identity records, and more. While HPE’s FIPS-validated FF1 implementation of FPE is not affected by the attack on FF3, it’s worth understanding a bit of the confusion about the NIST process impact.
For reference on HPE’s FPE position, refer to our last blog topic where we review HPE’s FIPS-validated format-preserving encryption, “At HPE, Strong AES FF1 Crypto and NIST Standards Matter”
So what is the current status and what’s new since the attack was announced?
- HPE customers using HPE SecureData based on FF1 encryption methods are not affected—it is still business as usual for the industry’s first FIPS-validated solution available for FPE that uses the robust FF1 method based on security proofs
- No new compromises in the cryptanalytic attack status since April 12 for FF3, and
- NIST has not yet determined next steps for FF3
With the above in mind, it’s worth looking at the reality of the current situation:
The NIST gold standard for security assurance helps determine vendor-independent trust
NIST standards and recommended best practices remain the benchmark of credible security assurance, both in federal markets as well as commercial. Notably, HPE continues to offer the only FIPS-validated FPE solution on the market with HPE SecureData with Hyper FPE based on FF1.
With heavy scrutiny and open challenges that are out in the public domain, security experts realize it’s more credible to be held subject to public peer review that helps remove the mystery of security compliance, than to simply take at face value vendors’ assurances . Trust should be earned and NIST remains the benchmark with public transparency in mind. Security vendors must welcome the critical public scrutiny of due process.
Alternatives to FPE may not be a relevant substitute
Traditional tokenization methods or AES encryption, for example, may not best offer the data masking flexibility, application usability with underpinned security, and similar values that make FPE best for data security applications where data in use protection is critical. And this assumes those alternative technologies are fully qualified as a starting point. Even so, relying on less flexible encryption approaches may not fit the needs of today’s modern application requirements, such as Big Data or IoT, where massive scale and usable data analytics are business concerns where FPE can help offer a perfect solution fit.
So with vendor credibility in mind, what should I consider to help ensure a trusted approach to FPE?
It’s important to understand how solutions are vetted to meet your needs vs. ambiguous claims that emerge:
Published methodology: Look for vendors willing to publish their methods for peer review and meet publicly-accepted standards. Peer review analysis helps ferret out methods and secrets, remove obfuscation, and avoid hiding behind claims, in hopes of achieving acceptable security methods.
Reliance on questionable expertise: Be leery if methods haven’t been analyzed by multiple, independent, expert third parties to help ensure credibility. Similarly, avoid methods that only examine security in terms of “brute force” risk. This is analogous to claiming, “the door can’t be broken,” when the lock itself is completely flawed. Whereas, NIST and similar industry standards bodies open up review to a wider audience who understand the credibility that these standards bodies have at stake.
The bottom line to remember is this—the cryptanalytic attack and review of FF3 is precisely welcoming of the wide scale and diverse scrutiny that ensures validation meets stringent security assurance criteria. Through process and procedure, potential technology adopters have an independent and trusted reference that supports maintaining a high bar for trustworthiness.
With HPE’s FF1 method, security was not compromised at the expense of performance shortcuts, as the design was prioritized to be secure against the variety of attacks, such as what compromised FF3. Nonetheless, public scrutiny is welcome, as it’s better to recognize exploits before they happen in the wild by more sinister actors.
While HPE customers using SecureData FPE solutions based on FF1 are not directly affected by the FF3 news, it’s easy to get caught up in the confusion of competing arguments and start to have doubts. We’re happy to continue the conversation by contacting us for more information to help guide you toward smarter, well-vetted, technology choices.
For more information, contact HPE Security – Data Security or your local HPE representative.