370
This free font can trick AI scrapers into swallowing gibberish instead of your content (www.techspot.com)
submitted 2 weeks ago by to c/technology@lemmy.world
51 comments fedilink hide all child comments
 

Most mass scrapers, on the other hand, simply grab the raw HTML underneath. ShieldFont exploits this difference through an automated process called OpenType glyph substitution.

That said, because the whole defense rests on scrapers reading code rather than screens, taking a screenshot of a shielded page and running OCR on the image can still recover the real words.

Screen readers used by blind readers also work from the code, so they read the decoys aloud. ShieldFont ships with a beta feature that provides those readers with the real text instead.

you are viewing a single comment's thread
view the rest of the comments
[–] [S] 77 points 2 weeks ago (1 child)

Per the article, yes:

Screen readers used by blind readers also work from the code, so they read the decoys aloud.

But:

ShieldFont ships with a beta feature that provides those readers with the real text instead.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 81 points 2 weeks ago (3 children)

    ShieldFont ships with a beta feature that provides those readers with the real text instead.

    AI scrappers will just pretend to be screen readers then.

    And if the approach becomes popular they will just OCR the text instead.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 6 child comments
    •
  • [–] 24 points 2 weeks ago* (1 child)

    Per the GitHub:

    Accessibility layer: <Shield> hides protected regions from assistive tech and ships an a11y prop that renders a real alternative beside them — mode: "text" puts your words in the page encrypted behind a time-lock puzzle the reader's browser opens (no link for a scraper to follow, no artifact for you to host), or mode: "audio" points at a recording you make. What remains: NVDA and JAWS verification (VoiceOver is done by hand, Windows is not), the focus indicator a sighted keyboard user loses to an invisible control, and the non-React tiers shipping none of it.

    In summary, they have two ways to get around this.

    • By default, blind people are fucked as usual. Most everything is unusable via screen reader just due to the way modern websites are designed.
    • Audio mode pre-records an audio file of your text, and plays that, bypassing the screen reader. This is extremely horrible UX for screen readers, for anything more complicated than a wall of text, but at least it's something.
    • Text mode will do some sort of computational challenge, using CPU on the blind person's machine for a few seconds, then revealing the original unmolested text. Assuming this works properly, because screen readers are extremely fragile beasts.
  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 22 points 2 weeks ago (1 child)

    Sounds like an overly convoluted way to do exactly what Anubis does...

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 3 points 2 weeks ago (1 child)

    Yeah, if you care about screen reader users, functionally it's just "Anubis but worse". Unfortunately, most people don't care, so for them and all visual users, it has the benefit of no additional "load time" computational check — the page appears instantly without the Anubis step. Though I'm not sure how long it takes the page to do the de-scrambling.

    The primary purpose of this project is to mutilate your HTML so bots can't scrape it, rather than preventing bot traffic in the first place. The screen reader stuff is a bolt-on.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 9 points 2 weeks ago (1 child)

    Unfortunately, most people don’t care

    There are laws about accessibility, at least for public websites, and probably for larger websites as they have such a large audience that disability can't be ignored (as much).

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 5 points 2 weeks ago (1 child)

    In my experience, those laws aren't followed much in reality. The biggest sites, sure, but most smaller sites are a mess under the hood. As long as something is "technically usable", even if very awkward and annoying, it's often hard to sue or anything like that. Maybe it's different in other countries though, I live in Australia.

    Next time you use a website, try only using your keyboard to navigate it and accomplish your task, and you'll see what I mean. It's a rough test of how much they value accessibility. For example, many websites disable the border around what you're highlighting, so you have no idea what the tab button is about to click. Or popup windows don't actually change your tab selection, so you have to tab through literally the entire page in the background to get to the popup window. Stuff like that; technically usable but annoying as hell.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 4 points 2 weeks ago* (last edited 2 weeks ago) (1 child)

    In the US, the Americans with Disabilities Act (ADA) facilitates tons of lawsuits against websites that are not accessible: https://www.rjo.com/publications/website-accessibility-lawsuits-continue-to-inundate-california-courts-despite-covid-19/

    Typically, in website accessibility lawsuits, a visually impaired Plaintiff will allege that a company’s website is inaccessible and incompatible with screen-reading software such that the individual is denied full and equal access to the company’s website in violation of the ADA and the Unruh Act. In these cases, a Plaintiff will seek statutory damages, injunctive relief, and attorney’s fees both individually and on behalf of a class of persons similarly situated. Plaintiffs have been testing this theory of ADA liability in courts across the country for years, with New York, Florida and California being the most active.

    Edit: here's another good link showing a graph of the number of ADA website accessibility lawsuits from 2016 to 2025: https://www.ecomback.com/annual-2025-ada-website-accessibility-lawsuit-report

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 2 weeks ago (1 child)

    That sounds nice. Wish we had an ADA, wouldn't even need to change the acronym lol

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 4 points 1 week ago (1 child)

    You do! It's the DDA (Disability Discrimination Act). A famous case is Maguire v SOCOG 2000 over the Olympic website. Also Coles got sued in 2014.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 1 week ago

    I always used to quietly rebel against my boss when I did web development work. I'd spend the time making the pages actually accessible to screen readers and tab navigation and such. Even fixing my coworkers' code to be compliant.

    I always thought it was a fun minigame, and it didn't even take that long. Little things like, icon buttons also need a text label, tabbing needs to follow a logical layout, shortcuts to jump to different parts of the page, actual tables and description lists.

    I used to get in a lot of trouble for "wasting time" because "no blind people are using our software." Years later an audit happens and my boss is shitting himself. We get a passing grade for accessibility. I never got any recognition for it, but he stopped yelling at me for wasting time after that and just let me do my thing.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 10 points 2 weeks ago

    AI scrappers will just pretend to be screen readers then

    There's a fair chance they're already doing that. It provides better insight on the content, less formatting to handle, and even visual stuff gets text alternatives.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 5 points 2 weeks ago

    It is also very easy to detect surprising text by looking at the perplexity levels by feeding it to a very small model. If there's a problem, OCR/'screen read' it instead..

  • permalink
  • fedilink
  • source
  • parent