Depositing in the electronic age

Last year, from time to time, a question pops up in my head that I still haven’t found the answer to: “How to realize the deposit of possession of some electronic information?”. That is, the statement of the problem is as follows: “the user of system X transmits a copy of some digital information Y, which he owns, to the depositing party D so that in the future he will be able to confirm to any third party the fact that he owned Y at the time of deposit T ". Purpose: mainly copyright, possibly preprints of some scientific research, in principle, any information in digitized form (even without D disclosing its essence for D - it is enough to file a crypto-resistant hash sum from it itself).


The options that are known to me today (all of them are essentially offline anyway):
- notary services (plus: undoubtedly advantageous in court, cons: 1) it is offline, 2) and in the notarial system in our country with all my respect for it there are facts of abuse), example: (no advertising, just the first link in Yandex) firm Seneschal;
- sending to yourself by mail a registered letter with a copy of information Y on some carrier, followed by keeping the envelope unopened - a stamp of the RF Post Office will confirm the date (minus: the proof is valid once, which was discussed in detail in this short form and comments to it);
- rent of a deposit cell in the bank specifically for this instance of Y - the date certifying factor will be the entry of access dates X to the cell, after placing Y in it, it is impossible to open the cell before the fact of the proceedings (minus the same one: one-time scheme).

Actually a question, to the discussion of which I invite all interested:
“A. Is it possible to develop an exclusively online scheme?
+ B. An even more popular option: ... with the impossibility of abuse by D? ".
')
In fact, we fall into the so-called region. "Cryptographic protocols." All that I have found is repelled by the postulate of unlimited trust in D on the part of all the participants in the system. The only light ray that I know is the protocol (for example, here ), which allows at least guaranteeing the range bounded by two adjacent in time (previous and subsequent) deposition facts. However, if the system works at half strength, this range can stretch up to say up to a week, which is probably already unacceptable in practice. If they fix me, I will be very happy.

But around us is the real world. There is no limitless trust in the real world, especially when the outcomes of court cases with millions of lawsuits can be at stake (if such a system is suddenly launched and “gaining momentum”), and there will not be. Perhaps the solution is that such giants of the IT industry act as a depositing party, that for them even millions of dollars would be less important than reputation (Google, Microsoft, etc.)?

Or is it possible to find an exclusively theoretical (“pure”) solution?

For example, creating a system of 3-5-20 independent companies, each of which constantly observes (saves in the archive) the published registers of the hash sums of the information deposited by "competing companies" with "mixing" in the signature algorithm of the sequence number in order to exclude the possibility abuse by the D publishing of the document in hindsight.

Perhaps this mechanism can be fairly organically “embedded” in various time-machine servers that store historical copies of various sites as of a given date?

Sorry, that messy. What is your opinion? Perhaps a perfectly perpendicular solution of this question is known? Why there are no similar services on the market?