Statement

Digital Access, Preservation & DMCA Rights

A tale of IFLA, censorship, and the fight for digital rights

What they got wrong...

About a year ago it became clear that IFLA websites were unreliable. Not just slightly unreliable, but there were periods of days, sometimes over a week, when they simply weren't accessible. People were unable to get information from the self-proclaimed Global Voice of Libraries, and that's a problem.

Whilst we heard a lot from them about "Security" and there are genuine security issues they haven't fixed, this was obviously not an issue of security but a failure of leadership and governance to capacity plan. Trying to cut resources under increasing demand.

As a voice IFLA is about communication. IFLA speaks for the field about access to information, yet its leadership was seemingly happy with poor or no access for significant periods of time. Which clearly harms IFLA's work.

What they got right...

IFLA has historically shared content using Creative commons. Specifically CC BY 4.0. This is a very permissive license. Anyone may:

Share — copy and redistribute the material in any medium or format for any purpose, even commercially.

Adapt — remix, transform, and build upon the material for any purpose, even commercially.

The licensor cannot revoke these freedoms as long as you follow the license terms.

IFLA explicitly says: Unless otherwise indicated, content shared on this website is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0), which means you are free to copy, distribute, transmit, translate, adapt and make commercial use of the work without asking permission, provided that any use is made with attribution to IFLA.

This is good. In fact, it's very good. IFLA shares the information but they do not control it. It helps IFLA meet its core values like ensuring equitable access to information, because even they could not censor it or its distribution. Arguably few restrictions on distribution is the point of CC BY 4.0 licensing. It allows IFLA to say things like:

Constructive participation and the development of democracy depend on satisfactory education as well as on free and unlimited access to knowledge, thought, culture and information. in the IFLA UNESCO Public Library Manifesto 2022.

And it allows them to make a myriad of statements on behalf of libraries such as:

In particular, we oppose unjustified restrictions on access to content online, or other distortions of information flows. We also argue for maximum transparency around the way that decisions are taken about what information is displayed, and how, as well as meaningful possibilities for internet users to choose how information about them is used.

in this statement, amongst many others.

When considering licensing we also need to keep a few other things in mind. For example, "moral rights". This covers things like the right to attribution and the right to object to distortion or mutiliation. Primarily we are talking about things where content is changed rather than reproduced as is.

Finally we should keep in mind there are some situations where IFLA cannot choose the CC-BY-4.0 license. This occurs, for example, where other licenses such as MIT or GPL require derivatives to be under the same or a similar license. For example, small changes to the CSS of a Wordpress theme would be a derivative GPL license rather than CC BY 4.0. But for our purposes here they are equally permissive.

I cannot state this strongly enough. IFLA's long standing policy of always using permissive licenses is highly commendable. Say what you like about them in other ways, but even the most controversial IFLA leadership till now has stuck to these principles.

A solution to the problem

CC BY 4.0 licensing provides a simple solution to IFLA's unreliable websites. As long as someone clearly notes the content is from IFLA, is licensed under CC BY, and follows the license terms, they are free to reproduce it to make it accessible.

This is called a mirror. If the original site doesn't work, you can use the mirror. That's completely legal, benign and helpful.

What it also enables is decreasing the load of the main site. By creating a mirror in a way that can handle far far more traffic and being permissive we can attract the high load but not entirely useful visitors such as rogue bots and scrapers to use the mirror. Allowing the main site to be more responsive to users visiting it to learn about libraries! Again, benign and helpful.

So I set up www.iflamirror.com with the express permission of their CC-BY licensing - which encourages reproduction. And it worked. So well in fact that IFLA followed later in using the same provider and principles.

It's important to state here that I lose money from this as I pay the costs. These mirrors produce no income whatsoever. I simply paid to ensure information was accessible using CC-BY, when IFLA's sites were not.

That's like paying so your neighbour's newspaper still gets delivered when they're clearly struggling. Everybody benefits.

Fixing privacy issues

IFLA collected some analytics data without permission. That represented a legal and ethical risk for me, so the mirror removed this tracking functionality.

In line with CC-BY-4.0 requirements, I added a notification that these trackers had been removed.

That's a legal requirement of the license that IFLA themselves chose. Whilst it may be uncomfortable for them to see it, it was necessary and proper under their terms. I don't think anybody would realistically expect me to risk being fined.

Digital Preservation

What began as a simple access solution became important for digital preservation. Reports showed that IFLA had been limiting crawling from Internet Archive and that historical IFLA sites were disappearing from the internet.

IFLA's digital history was being silently erased, despite the organization's public commitment to preservation.

Thanks to the CC-BY 4.0 license, these sites could be preserved. I added subsites to the mirror, ensuring accessibility and that knowledge remained available. In line with IFLA's core values and principles.

This makes it obvious just how much IFLA has been silently deleting. In some fields preserving them might be controversial. But not in this field.

In fact, IFLA made this statement in August 2025 arguing for:

And whilst CC-BY means those rights already existed here. I'm sure it is hard to miss the point that the right to do this is woven into the fabric of IFLA's being itself. Across many many documents IFLA argues that this is a right, to the extent that even if CC-BY didn't exist there would be no trouble arguing permission is implied. The rights to preserve digital history are woven into its very existence.

Supporting IFLA’s Principles

The mirrors did exactly what a supporter of IFLA’s mission would expect: they preserved access, ensured information was not lost, and followed the terms of the CC-BY 4.0 license. No profit was made, no content was altered (with the exception of illegal tracking), and knowledge remained available for anyone who needed it.

In short, these mirrors aligned with IFLA’s own commitments: digital preservation, equitable access to information, and the free flow of knowledge. They reinforced the principles IFLA publicly promotes, exactly as the organization would encourage in theory.

From this perspective, one might naturally assume that IFLA leadership would see this work as helpful, rather than a problem. I did. Unfortunately, the next section shows that assumption would be completely unexpected.

8 Emails from Cloudflare

On 3rd March this year, I received 8 emails from Cloudflare. Unfortunately these would go unnoticed for some time as I was indisposed.

These attacked the 4 of the mirror sites, with 1 being the main one.

The sheer volume of these requests would've hinted that something entirely disproportionate was underway from a previously silent IFLA leadership, had I been able to read them!

Trademarks

IFLA's Secretary General, Sharon Memis, filed four trademark claims. Each claim was essentially identical:

Original Work Description:
Trademarked Symbol: IFLA
Registration Number: 1494498
Registration Office: Benelux Office of Intellectual Property (BOIP)

It is of concern to me when an organization such as IFLA that expends so many resources on promoting the ideas of access to information, preservation, and equality, decides that the answer is lawfare. In this case abuse of trademark law to try to force entire sites and information off-line.

Sharon Memis' predecessors worked hard to enshrine the importance of these things. This is after all the overwhelming majority of Article 2 of IFLA's statutes. IFLA repeatedly describes this as a "right". And it is the reason why IFLA has, for a long time, applied CC-BY to its work.

Had Memis expressed any interest in following the work of her predecessors, and actually improving access and preservation. I'd have gladly handed these sites and historical data over.

Instead we have a situation that is uncomfortable for me, and should be uncomfortable for the field and IFLA. We have an organization advocating for the right to do this, where its leader is trying to backchannel legal technicalities to add a silent addendum of: "but not if it is us and information we are removing".

On the face of it, the choice I am given is to takedown the sites - removing information from the internet entirely. I find this argument very weak, but I have neither the time nor will to argue it. So I acquiesce to claims related specifically to the trademark as described in the registration. This has been replaced with an alternative indicating the reason for removal.

DMCA Requests

58 minutes after the trademark claims, I received 4 separate DMCA takedown requests from Cloudflare. Each was essentially the same, so we will examine one.

A DMCA request is supposed to identify specific copyright violations and provide evidence. It must be factual, not defamatory, and actionable. What I received instead was something very different. Let's take one example: Sharon Memis filed a DMCA notice claiming that the mirror site was an "unauthorized 1:1 copy" and "fraudulent," among other things.

In my opinion these are clearly quickly generated AI slop. There are just too many signs that is the case. If I had to guess this is Copilot, possibly Gemini.

As with most AI stuff they make many claims, some of them wild, but completely unsubstantiated with any evidence whatsoever.

It claims "the above website on the 'iflamirror.com' domain proxied through Cloudflare", but it is not proxied.

It claims "of an official International Federation of Library Associations and Institutions (IFLA) copyrighted website.". Yet it absolutely is authorised - for anyone. In IFLA's words: Unless otherwise indicated, content shared on this website is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0), which means you are free to copy, distribute, transmit, translate, adapt and make commercial use of the work without asking permission, provided that any use is made with attribution to IFLA.. CC-BY-4.0 does not need permission (as the IFLA statement says) and is non-revocable. Memis cannot simply decide she wants to disappear a site of information from the internet contrary to the values of IFLA, and the hard work of predecessors, by unilaterally revoking something she cannot revoke! Again beyond the fact that it is licensed, IFLA time and time again argues and advocates this as a right. They don't argue an exception clause for IFLA. They don't argue that Memis is the arbiter of when this does and doesn't apply. It's clear here she thinks she is backchanneling rather than making public claims. When she thinks she is not being watched, her actions and her arguments are very different to her words and IFLA's stated values!

She claims "First of all, not all of the intellectual content on this “mirror” site is licensed CCBY4.0, but the mirror site has reproduced these nonetheless.". Under the DMCA she must provide links. That avoids the situation that existed where I looked for the "unless otherwise stated"s. Hint - there were none. That's why the procedure of requests is important. Baseless claims waste the time of those on the receiving end (3 days so far). Given IFLA policy is, and has been for a long time, to use CC-BY it is also highly unlikely that Sharon Memis would be the copyright holder in any exclusions to be able to file a DMCA complaint.

Whilst claiming the site is a 1:1 copy she claims that it is an "attack on the moral rights". Yet seems oblivious to the fact that CC-BY licensed information she claims is repeated 1:1 cannot possibly be attack on the moral rights of anyone, unless their original was also an attack on the moral rights. By writing content on IFLA then given IFLA's statements on preservation and describing it as a "right", the expectation has to be that it will be around forever. Licensing under CC-BY accepts and encourages that that may be a copy. A mirror archiving content preserves "rights".

The mirrors according to Memis are "fraudulent" and "deceptively reproduced by an unaffiliated individual or entity". Yet despite the dangerous nature of making such claims in a legal document subject to a penalty of perjury and from a jurisdiction where these claims could be considered criminal, she provides absolutely zero explanation or justification for these whatsoever. These statements are reckless and unsubstantiated. Think back to IFLA's statements on the importance of protecting digital history - the "right". If it is IFLA deleting digital history then what actually happens is the IFLA Secretary General defames you thinking they're on a backchannel rather than public.

Worrying for those in the library field working in or connected to copyright, is the only point that Memis makes that still doesn't point to specifics but we can at least quantify:

Moreover, the website’s structure and design, which has been completely copied (including stylesheets), is not subject to the CCBY4.0 (content) license. This constitutes copyright infringement.

Why this should worry you is the website's structure, design, and stylesheets are made up of GPL licensed Wordpress, in this particular case the GPL licensed Wordpress Twenty Twelve Theme, and IFLA changing a few settings in the control panel to make it green.

Let's think on that for a second! The Secretary General of the international organization that helps and advises libraries on things like copyright filed a DMCA claim under penalty of perjury where the only remotely quantifable claim is that she owns the copyright to GPL licensed software.

I genuinely do not think I need to say more.

The Real Risk to IFLA

Memis talks about reputational risk a lot. So I think, when we come to a juncture where she has thought she could use trademark law silently to disappear IFLA sites from the internet forever. And where she has claimed IFLA hold the copyright to the largest open sourced GPL licensed website management software on the planet. That we should step back and consider what the real risk to IFLA is here.

Memis clearly thinks she is backchanneling rather than filing public legal notices. That's the only thing that could explain the blatant reckless and unqualified statements, the blatant lawfare, and the blatantly different position to IFLA's public statements. But that's not an excuse. It's merely a naive revelation of true nature and risk.

Every DMCA notice is recorded in databases such as Lumen. Researchers use that to study copyright enforcement, censorship, and abuse. I'd be pretty sure Memis didn't know that. But before trying to abuse copyright process to remove sites from the internet with possibly the most ludicrous claim she could've made, would it not make sense to look it up?

That's her name, IFLA's name, a record of a ludicrous DMCA claim, and a digital preservation of the site so everyone can see it's just Wordpress Twenty Twelve - preserved for ever. That's the addendum to every time IFLA advocates that digital preservation or digital access are rights. A big virtual "* but not if IFLA wants to remove information".

Isn't that the greater risk to reputation, credibility, and influence? That Memis chose to be heavy handed to strongarm information off the internet despite IFLA's values and principles and ends up publicly claiming copyrights to Wordpress components?

When IFLA talks about access to information, preservation, and equality isn't the risk to reputation that IFLA leadership looks like they think that applies to everybody else and not them. Now Memis doesn't need to like the mirror and archive sites. Some people will, some not. But like here, they can have a legal right to exist. What changes things is what lengths somebody is willing to go to to try to force their will against the rights of others. That's where the risk arises.

It seems the risk that IFLA's leadership is currently willing to take is to have a public record of their actions contrary to IFLA's stated policies?

And I have to make the connection. Isn't the real risk here that IFLA has failed to change its culture to follow it's public statements, values, and principles, and when they think they are backchanneling IFLA leadership are still aggressively controlling of others?

Oops

It is standard practice for organizations receiving DMCA requests to pass them on and set time limits for reply. If they don't get a reply then simply take the content down. I was unable to reply.

There's a balance issue here. While modern times mean a complainant can produce an AI slop claim on the most ludicrous of grounds in a matter of minutes, the respondent is given 48 hours to 72 hours to reply to something that often has no actual points or have their site taken down. I would say that should be something for IFLA to take up, but they're part of the problem right now!

It's a system that isn't based on rights, or right or wrong, but simply the amount of time a respondent has available. What should be hard, what should take effort, and solid reasoning, to remove information, is as simple as ludicrously claiming you have the copyrights to Wordpress.

And who do we blame? Do we blame the hosts who are simply doing what they have to to avoid risk? Do we blame the respondents for being slow? The lawmakers who make manipulable systems? Or those who manipulate them?

I did not respond in 48 or 72 hours. In fact, it was much longer. So some parts got blocked because I didn't respond in time.

Let's be realistic about what happens from here. I've emailed Cloudflare, explained I could not respond at the time and asked for reinstatement on the grounds there are no answerable claims in Memis' DMCA requests.

But, like I said, it is not a process based on right or wrong. It's a process based on risk. I represent no risk to Cloudflare at all, rolling back the takedown would. There is absolutely no incentive for a commercial company to do that at all, but in fact every disincentive.

How hard do I try to resurrect them?

To be honest I've been unavailable a lot longer than 72 hours. A month. And I sit here right now a bit in shock.

Nobody had problems nearly a year ago when I set them up. Nobody complained despite plenty of IFLA HQ visitors.

The mirrors aren't promoted, they don't appear to get many links and the ones that do exist tend to be relying on it for important historical context that IFLA has chosen to simply erase from the internet despite its values and principles. Functioning exactly as intended then.

Nobody is forced to like it or not. Just to recognise the right to preserve access and history that IFLA has historically granted and continues to grant despite Memis' actions.

But there's a question of balance.

Just not liking that you cannot delete stuff from the internet is not a reason to publicly call someone fraudulent

Just not liking that you cannot delete stuff from the internet is not a reason to publicly call someone deceptive.

Just not liking that you cannot delete stuff from the internet is not a reason to abuse trademark law to try to take it down.

Just not liking that you cannot delete stuff from the internet is not a reason to make ludicrous copyright claims to also try to take it down.

That's unreasonable. It's disproportionate. And IFLA have known that there is public record of this for at least a month and stayed silent.

Are we adults here disagreeing on IFLA's definition and responsibilities for access and preservation? Or something else entirely?

Cloudflare acted narrowly, given extraordinarily narrow complaints, and disabled 4 single pages. The rest work fine.

In a strange way, that outcome says everything. The majority of the content is preserved, accessible, and lawful. It continues to exist.

From a preservation perspective, the scars left by the blocked pages now tell part of the story. They are marks of what happened. Evidence of a story that people can follow with the Lumen database and the mirror itself.

It's amazing how, even in the most testing of circumstances, things often seem to find their own balance.

But realistically can I monitor them every 72 hours to prevent recurrence? Probably not.

And whilst my original vision was clear, well defined, and helpful. I also did not expect this experience. To put this as delicately as I can - I don't feel net joy or fairness!

I don't see any need to try to resurrect anything. The small scars are part of a verifiable history. And the rest can stand on its own.

Claiming copyright on a Wordpress theme indeed

Am I going to wake up now? Surreal.