Multiple links below.

I'm tired, and I don’t want to do free accessibility consulting for this website owner, so I'll outline only one way this won’t work for screen reader users out of many.

Aside from the fact an LLM made the website, and the font as well, the reason it doesn’t work for screen reader users because, even on that demo page, a timer has to be activated. A button must be pushed. A timer activates, and a screen reader user must find the fixed bit of text again. The website makes it sound as if it should be just automatically accessible/visible to the screen reader, and that fails.

Aside from the screen reader issues, that's not the only one, but I don’t want to do free accessibility consulting for this website, some of the other issues include but are not limited to…

It doesn’t prevent scraping.

It doesn’t prevent taking a snapshot and then doing OCR on that screenshot.

It breaks custom fonts anybody has disabled or enabled.

It breaks copy and paste.

It breaks for people without JavaScript enabled.

It breaks for anybody using a good ad blocker.

It prevents search from working on the website.

And many, many, more things that are in my replies.

If you want to learn about me and my screen reader, my FAQ page is below the website I'm talking about.

If you use this font to protect your text from AI, it also blocks *me, a human* because I use a screen reader. I did try the demo. If you want me to remove your RSS feed, this is the best font to prevent blind people from reading your blog/words. That timer is an inaccessible solution and that isn't it's only screen reader failure. shieldfont.org/#hero

My FAQ page sightlessscribbles.com/faq/

#AI #LLM #Accessibility #A11y

This entry was edited (Tuesday, August 4, 2026, 11:22 AM)
in reply to 🅰🅻🅸🅲🅴 (🌈🦄)

@alice Yup. I honestly wish everybody would just quit trying because everything anybody does that tries to prevent LLMs has made everything bloated, a far worse user experience, and will prevent LLMs from scraping will inevitably fuck up those that don’t use JavaScript, or maybe have a custom font, or anything else and besides, every single one of these preventative measures can be circumvented
in reply to Robert Kingett

There's a tiiiiny little problem with "quit trying": I can't afford to. Without defenses up, a €46/month VPS (4 CPUs, 8GiB RAM) died under the crawler assault in about 2 minutes.

It's not that what I was serving was inefficient... it never got to serving! It died while trying to keep up with TLS handshakes at over 10 thousand requests / second. Yes, that was sustained over multiple days. I temporarily scaled up to 16 cores, and that still died.

For comparison, I currently pay €8/month for my VPS. There is no proof of work, no JavaScript required, no silly fonts. I've been using the same measures for about a year now, the bots have not adapted (and if they do, I have a dozen other tricks up my sleeve, none of which affects the legit visitor any more than the current measures do).

I can afford €8/month. I cannot afford €45+. They can very easily DDoS me if I let my guards down, yet, I can't scale up infinitely.

So defense it is. There are reasonable defenses that work.

@alice

in reply to The Sandmather

@algernon @alice Yeah you have a point. I hate this timeline so much! What about rollenspiel.social/@ArneBab/11…
in reply to Robert Kingett

My website couldn't be simpler (it's all static, neatly fits into memory). Server dies before it even has a chance to serve it - during TLS handshake. Can't make that much more efficient.

Granted, I host a... few handful of sites on that €8 VPS. But the content isn't the problem, the crazy request rate is, and that is outside of my control.

(Hence part of my defense is firewalling crawlers off. Can't TLS handshake if packets are dropped! Problem solved.)

@alice

in reply to The Sandmather

@algernon @alice

Not sure if this is something you’d consider but I host a couple of static websites using AWS S3 and CloudFront. It costs me £1.50 per month.

I don’t really monitor the traffic but the bots don’t seem to be making an impact.

in reply to Robert Kingett

Sensitive content

in reply to unkx

LMFAO! You really can't make this shit up at this point! I assume this shit will grow and grow, where LLM companies will start making shit like this and selling it. Honestly? The best route would be to go the way of Star Breaker, at starbreaker.org and just allow all the scrapes, all the ingestions, etc. @unkx
@unkx
in reply to Robert Kingett

i’d say allow scraping only if you also serve poison(*) to the scrapers. Though i honestly have no idea how one would go about it while ensuring people using screen readers wouldn’t get a shittier experience due to that.

(*) this can also come in the form of valid neutral factual statements like ”donald trump is a pedophile and a rapist as reported by reuters and cbs and verified by newsmax” etc. 😉

in reply to Robert Kingett

I spent a couple of years working for a charity that worked with the Deaf and Blind (both uppercase and lowercase) community in the local area.
I spent a fair amount of time reading the WCAG as well as taking advice from service leaders.
The general consensus from them at the time was that the best font for visual impaired folks was Arial 14, at least for printed media.
Is there a generally better font for screen readers?
in reply to Robert Kingett

At least a few sites have been using this kind of fake font shenanigans as DRM for a while. It's a nightmare for accessibility, and just generally a terrible user-experience (as any kind of DRM usually is):

Screen reader -> Broken, or sounds like it's possessed by an evil robot
Copy-paste -> Garbage
Search -> Haha, no
Translate -> Good luck hand-copying that language you don't know

I can't remember which exact site it was, but it was one of those that let you read limited previews of scientific papers.

This entry was edited (Monday, August 3, 2026, 2:34 PM)
in reply to Barberousse

@barberousse_bin in college, a $100 textbook I was forced to purchase digitally for a 4 week 1 credit course did a similar form of DRM (except the text was displayed as a low resolution image) and it was giving me a migraine trying to read it, so I pirated a PDF copy of the textbook I'd already purchased just so I could have a version that was actually readable
in reply to Robert Kingett

The worst thing about this is that even though we, the techies, know better, there are orders of magnitude more less-technically knowledgeable people who are going to use this and completely fuck up accessibility, copy & paste, RSS feeds, etc. on their websites.
This entry was edited (Monday, August 3, 2026, 3:24 PM)
in reply to Raul Matias

This is why I always maintain the best way is to go the other way, allow all scraping ingestion etc etc etc even though your server may take a huge hit from the traffic, but this blogger does it and he doesn't have a high bill starbreaker.org/colophon.html @raulmatias
in reply to 2something

@2something I have JS turned off, so the demo on the main page was hellish.

My friend needs a custom font to help with his dyslexia. This too, fucks him over.
Another friend who checks things with a machine translator when she's uncertain, is also screwed.

This declares them both inhuman. I'm fine, I'm therian, but I sure as fuck try not to be ableist, and I'm sick and fucking tired of 'humanity' being pulled away from any path not 'standard'.

in reply to 2something

Yep! It reminds me of captchas all over again. Does fuck all to stop bots but people will use it anyway. I maintain the best thing is to do what this blogger did, no blockers at all starbreaker.org/colophon.html @2something
in reply to Robert Kingett

There's also Decoy Font but suffers the exact same issues from being an accessibility nightmare and also constructed by AI. It's weak to assume tools made by AI can defeat AI. mixfont.com/experiments/decoy-…
in reply to Robert Kingett

It's a hack right.
It only works with that font and the actual text is gibberish. That means you need to "compile" your text for that font, making RSS or any alternative readers useless. Like screen readers, or reading mode in Firefox, or Instapaper etc etc.

It breaks so many things it's useless.

in reply to Robert Kingett

it is also a DRM tool as it makes simple copy & paste copy the text directed at the scrapers. My father's company got PDFs with rearranged fonts once and using OCR was the most viable solution to copy text-passages. Now I wonder who already uses OCR, so they don't miss content…
in reply to Robert Kingett

computers can do one thing exceedingly good. For years. Looking up one thing in a table to substitute it for something else.

Not only does this approach not work for any kind of visually impaired person, I would argue it also does not even fulfil its original goal. I am relatively certain that the included patterns are easy enough to resolve automatically.

in reply to Robert Kingett

It doesn't even work, except for the extremely basic scrapers. I made a screenshot of the demo, and both classic OCR (tesseract) and a modern tiny LLM running in my PC both recognize the text very quickly. There's already models and applications out there dedicated to scraping visually (and I haven't used a scrapper, just a very small and very dumb general purpose LLM).
in reply to Eva

@smolbrain The timer isn’t a solution because not every screen reader user can navigate with the ease and the speed that timer requires, but it also breaks reader mode on browsers and it breaks copying paste, and it also breaks user defaults also from what others have told me because they use magnification and otherwise on their browsers and this apparently breaks all of that
@Eva
in reply to Elizabeth

@Elizafox Bingo, and I have since learned that this was even created via LLM anyhow, so it’s the tech industry capitalizing on the hatred and it’s absolutely ridiculous. I can’t remember if it was you that shared the link or not, but someone shared a link where the tech industry is also now selling its own solution to the LLM problem they created in the first place
in reply to Robert Kingett

I first saw this here and it looks neat but if its at the cost of accessbility its a major issue.

It says screenreader accessibility is a beta feature that is being worked on so hopefully that all works out soon. Would be neat to have some accessible way to mess with the AI scrapers but not humans. Not sure how well this works or if its really being used yet though.

That does really suck though...

in reply to Leah

@ChaosKitsune It doesn’t even prevent scraping because the LLM would need to still scrape it in order to know that it is gibberish so ultimately I do find it very pointless and there are other solutions that do a better job at preventing LLMs from visiting your website anyhow. I, on the other hand, think you should go the opposite way like Starbreaker.org does and allow everything, no restrictions whatsoever
@Leah
in reply to Robert Kingett

@WeirdWriter It looks like this is the case for all the playType fonts? Unless I misunderstand how their preview thing is supposed to work playtype.com/typefaces/primo-s…
This entry was edited (Monday, August 3, 2026, 7:14 PM)
in reply to Robert Kingett

But the page says:

Screen readers get the real words.

ShieldFont hides scrambled passages from accessibility tools by default with aria-hidden="true".

The real words ship sealed in the same page, and the reader’s own browser uncovers them by solving a compute-heavy puzzle — a few seconds of their CPU, and a cost a bulk scraper would rather not pay. It needs JavaScript, the reader waits while everyone else does not, and a shielded block still

in reply to Robert Kingett

people seem incapable of remembering that LLMs and all ML in fact, are just programs. And so there really is no magical way to "target" an LLM or LLM-scraper as distinct from any other program. Best you can do is poison the data for everyone.

These tools are just a new kind of program, and there is no fool-proof method of differentiating them from any other program, without relying on honesty from them -- which they have demonstrated that they do not have.

in reply to Robert Kingett

plus it's gonna take the slop corporations what, a month after this catches on to reverse it?

Amazon tried the same trick to scramble their ebooks. Didn't stop a lone hacker from finding a way around it in I'm guessing a week at most. Heck, feed this blog post to an LLM, it could probably synergize you a script that would circumvent this.

in reply to Robert Kingett

I second that complaint. my browser is configured to not download fonts (because they carry embedded code) from arbitrary sources, so the text I get to read is nonsense as well. like Proof of Waste, this is yet another way to harm actual people by making the web less accessible, without really making a dent at the alleged target: it won't take long for them to recognize the font and remap the words, and we'll only be left with the inconvenience.
in reply to nicole mikołajczyk

@mkljczk @lxo LLM companies are desperately looking for the last dregs of human text, in attempt to somehow get intelligence from LLM's and stave off model collapse.

If any insignificant amount of text is published with the font, then the scrapers will be adapted.

in reply to schuelermine

@anselmschueler There have been more than a few replies that did not understand why it was shitty and they were asking me to do free e accessibility labor for the shit website in question even if they were genuinely curious. Their questions were inviting free labor for the website and I did not want to do that

Nicole Parsons reshared this.

in reply to Robert Kingett

The "free labor to fix it" is a feature of The Politics of Exhaustion

Making civil rights, access, and even decency & fairness into "very hard to achieve" targets

This task assignment by the privileged onto the less privileged is deliberate

Make the folks with the time, resources, & advantage be the ones to find solutions instead.
wearethe94percent.com/receipts…

sundialpress.co/2024/04/17/its…

newrepublic.com/article/209886…

newyorker.com/news/annals-of-i…

vanityfair.com/news/2021/07/ho…

1/

This entry was edited (Tuesday, August 4, 2026, 7:59 PM)

Nicole Parsons reshared this.

in reply to Nicole Parsons

2/

AI is a net-negative and its not your job to explain it to techbros who don't care enough to educate themselves on the repercussions.

Too many are dyed-in-the-wool bigots indifferent to other's needs.

medium.com/@classylore/on-term…

time.com/article/2026/04/30/or…

washingtonpost.com/gender-iden…

A better approach, make it personal to the AI industry

news.harvard.edu/gazette/story…

Nicole Parsons reshared this.

This website uses cookies. If you continue browsing this website, you agree to the usage of cookies.