Sustainability of ZBmath unrestricted access ZBmath (formerly Zentralblatt für Mathematik) will become "Open access" in 2021 (see for instance at EMS site and at FIZ Karlsruhe site). It used to be under paywall, although some partial access was allowed.
My question (which was suggested in the comments of this answer to Is a free alternative to MathSciNet possible?):

What are guarantees (technical and/or legal) that the reviews will remain in unrestricted universal access?

(Concretely, I'd be reluctant to write reviews if ZBmath has the possibility to suddenly turn back to partial paywall in 2030 for whatever reasons, especially if this paywall affects searches for reviews made during the so-called open-access period.)
 A: Thanks a lot for the question! On behalf of Klaus Hulek (as zbMATH Editor-in-Chief) we can confirm that it is correct that we will go Open Access as of 1st January 2021. This will mean that the database is freely accessible by everybody worldwide. It also means that most of the data will become open via a CC-BY-SA license. Such a license cannot legally be revoked. This applies in particular to the reviews written by our reviewers (only excluding the very rare case of reviews where the copyright is owned by a third party, which amounts to less than 0.5% of the reviews). The other main exception is third-party content such as author summaries, which may be provided under a different license. We will further have an API which will allow users to download parts (especially, the CC-BY-SA-licensed parts) of the database.
A: On the legal side, I expect that this will depend quite strongly on what precise kind of open-access policies they will implement. If "open access" means just "no payment to read" (but all reproduction rights remain reserved), then I suppose they could simply change the policy again at a later point in time. On the other hand, if "open access" means that articles are released under e.g. a Creative Commons license, then anyone can upload those articles to repositories that will preserve them indefinitely, and since the CC license once granted cannot be revoked, there would be no legal way of suppressing the free distribution of those articles in the future.
(Caveat: I'm not a lawyer, and this is not legal advice.)
On the technical side, I don't see how there could be a viable guarantee that access isn't restricted at a later time.
So in summary: in my opinion, "open access" is only truly open access if the contents are licensed appropriately to enable free distribution.
