On Tuesday of this week (25 February 2025) every major print newspaper in the UK carried an identical front page under the heading “Make it FAIR”. They did this in order to raise awareness of the impending legislation (currently in “consultation”) intended to change the law in the UK to favour big tech platforms in order that they may use British creative’s copyrighted material to train their AI models without permission or compensatory payment.
This is plain stupid – particularly at a time when US politics is highly volatile and in thrall to the very same big tech companies which own these emergent AI technologies.
The print media coverage pointed to a campaign letter asking readers to contact their MPs to urge the UK Government to see some sense and block this pillage of creative content. As a (small) creative content author, albeit one who is not concerned about personal monetary reward, I read the campaign letter and decided that I should lobby my MP, Ben Goldsborough, in line with the campaigns aims.
What follows is my response to my MP’s boilerplate reply to my initial email.
On Wed, 26 Feb 2025 09:25:36 +0000 (UTC)
ben.goldsborough.mp@parliament.uk allegedly wrote:>
> Thank you for contacting me about creative industries and the
> application of copyright laws to artificial intelligence (AI).
>
[ some deletia ]Mr Goldsborough
> I support right holders to exercise control over the use of their
> content in the training of AI models and their ability to be
> remunerated for their work. Writers, musicians, artists and other
> creatives should have the ability to know and control the use of
> their content, as well as the ability to seek licensing deals and
> fair payment.I am glad to read that you support rights holder’s rights to exercise
control over the use of their content. I am, however, not at all
encouraged by your (DSIT “boilerplate”) response to my initial
email.> The application of UK copyright law to the training of AI models is
> currently disputed and these are the challenges right holders face
> today. Right holders are often unable in practice to control how
> copyright in their work is used or to gain remuneration for their
> works, this is often particularly true for new or solo artists.Indeed, it is often very difficult for creative rights holders to
control access to, and use of, their copyrighted content and I am
only too well aware of the impact of the major tech media companies
(such as Napster, Spotify et al) on the fair remuneration of new or solo
artists. This has already gone too far in favour of the tech companies
and against the best interests of the content creators. To now consider
permitting the major AI tech companies (who are predominantly US based
and owned) full access to copyrighted UK creative’s content simply
because it is available on-line to be data mined goes way further than
should be premitted. This is particularly sensitive at a time when US
politics is dominated by the interests of global behemoths such as
Google, Meta, X etc. The UK should stand firm against this attempt at
cultural “burglary” (which is all that it is) and I hope that you, as my
Parliamentary representative, will add your voice against any such
attempt.> This situation is hampering innovation and growth in both our creative
> and AI sectors. The Government is therefore consulting on plans to
> provide legal certainty to both sectors. I would encourage you to
> contribute your views to the consultation.By writing to you in this manner I am indeed doing my bit to contribute
to the consultation. As my representative, you are in a far better
position than am I alone to influence the outcome of this
“consultation”.
>
> The Government’s proposals include a mechanism for rights
> holders to reserve their rights, allowing creators to license and be
> remunerated for the use of their work in AI training. Under this
> proposal, right holders can expressly reserve their rights and AI
> developers are only allowed to train on materials in which right
> holders did not reserve their rights. This approach of a data mining
> exception allows right holders to reserve their rights underpinned by
> transparency and removes legal ambiguity.Forgive me if I say that I do not believe this. Nor do I see it as
technically possible. For background, I should say that I am (or rather
was in my professional life) an IT technician. I worked in system and
network administration (in Government I might add). Further, I have for
some 18 or more years run and edited my own technical blog (at
baldric.net). That blog is small, niche and largely technical in nature
and devoted to open source software – though I do sometimes stray into
non-technical topics. It does not have a massive readership, getting
around 100,000 to 170,000 hits per month from about 10,000 to 13,000
unique sites, but I do hold monthly statistics going back some long
time.The statistics I hold tell me quite a lot about my “visitors”.
Without going into unnecessary technical detail you simply need
to understand that every time a human being, or ‘bot, accesses a
website it leaves a very distinct “signature”. This signature includes
details of the “user agent” accessing the site. For example, if you
use Google’s chrome browser on your office PC running MS Windows, it
will leave a very different signature to that seen if you later access
the same site using Apple’s browser from your iphone. Bots (such as
those run by Google to index websites for its search engine) in turn
leave a very distinct, but different, signature.For some time, the predominant signatures in my statistics (in terms of
most often seen) are those of search engines. I expect this because
unless search engines can access and index a site, it cannot show up in
on-line searches. I don’t like the fact that random strangers suck the
content from my site, but I only apply a limited copyright to my
material (see https://creativecommons.org/licenses/by-nc-sa/3.0/) and I
permit search engines for the very simple pragmatic reason that unless
I do permit them access, I am limiting my readership.However, a perusal of my latest statistics shows a worrying trend.
Along with the (expected) search engines I see the signatures of AI
‘bots. For example, in february of this year, of the 137379 hits on my
site, fully 9356, or nearly 7%, came from openai.com/gptbot making that
robot the top user of my site for this month. Similar levels of activity
are visible over the last six months where the gptbot is usually in the
top five of “users” on my site.Now, there are technical limitations a web site owner can apply to
restrict such activity (the “robots.txt” standard is the most popular)
but such approaches have severe failings. Firstly, even if correctly
applied (so that I say “disallow the following signature”) the robot
might ignore that instruction. Or it could change its signature to
match one that I do allow (trivially easy). But that is all besides the
point. I should NOT have to erect a sign on my website saying
effectively “thieves not allowed”. A shopkeeper doesn’t have to do
this, it is simply expected and covered in law.> The Government recognises that for rights reservation and measures on
> transparency to work in practice, there will need to be appropriate
> support to encourage widespread adoption and compliance. The
> Government is keen to ensure that rights holders understand how and
> where their content is used, while also ensuring any measures are not
> disproportionate for small businesses and individuals.Exactly how will the UK government enforce compliance of restrictions
on copyright theft on an egomaniacal lunatic like Musk?> I would like to assure you that this Government stands steadfast
> behind our world-class creative and media industries that add so much
> to our cultural and economic life.I’ll believe that when I see some push back on the tech companies.
Best
Mick Morgan
———————————————————————
Mick Morgan
gpg fingerprint: FC23 3338 F664 5E66 876B 72C0 0A1F E60B 5BAD D312
blog: baldric.net
———————————————————————