Thread by @jimkillock

A few thoughts on the #OnlineHarms White Paper. It has a "vision" of a "A free, open and secure internet [and] Freedom of expression online" "where companies take effective steps to keep their users safe".

In 102 pages how free expression will be protected is not explained.

There is a list of harms: "Harms with a clear definition" and "Harms with a less clear definition". These are roughly "illegal" versus "legal but unwanted". (p31)

Content may be legal and harmful yet will remain legal to publish unless you are a platform. Why?

Probably because it's very difficult to establish the harm from the content, or it would be contentious to ban that content.

This should signal the complexity and difficulty in balancing free expression and content takedown. This is not discussed.

Basically it is all down to the Regulator, who will be told to "particularly mindful to not infringe privacy and freedom of expression." Well, ok, good. However …

… the Regulator will be expected to ensure companies make "Proactive use of technological tools, where appropriate, to identify, flag, block or remove illegal or harmful content."

… which has obvious free expression and privacy issues.

It's ok tho, because it's essential that "users have an effective route of appeal".

Reliance on users to sort of out the mess caused by automated tools reminds me a lot of blocked.org.uk for instance, or of DCMA copyright takedowns.

We should be very worried.

If you want a quick summary of the main points, here's our Wiki page: wiki.openrightsgroup.org/wiki/Online_Ha…

So far I've not found anything explaining the evidence threshold for #onlineharms to require action. There are some reasonable remarks about dealing with the "most serious harms". But what is the threshold for action?

This really matters. If popular opinion deems that a particular kind of content is harmful, and should be removed, does the regulator act? What kind of evidence of harm is required?

Establishing a relationship between harm and content is in practice incredibly difficult. Assumptions abound.

If the evidence standard is low, we get over-reaction. If it is set reasonably, it may be impossible for the regulator to require action despite public demand.

The #OnlineHarms White Paper also seeks to protect vulnerable groups in particular, a worthy goal. However, this is difficult to reconcile with the ambition for proportionality. Vulnerable groups need a lot of protection: the rest of us will suffer from it.

My overall impression of the #OnlineHarms White Paper is that it is raising public expectations that cannot be fulfilled.

Sound familiar?

That is a long standing principle of UK Internet policy. Government drives public expectations, and the policy that comes fails to deliver.

Will the #WhitePaper be another example? Perhaps, but whatever the result, at least the government can blame the new Regulator.

Summary: the #onlineharms proposal raises some very serious questions.

» Content that is legal can be de facto banned.
» Content can be removed without due process
» 'Prior restraint' on publishing can be established
» Little evidence of harm may be employed