this is an encryption algorithm that replaces characters by others, using a dedicated font. The article explains that the idea is to prevent LLM to scrape the content.

Apparently, the person behind this website is an ableist person because they know they're sacrifying accessibility and that it doesn't work with screen readers.

That's real shit.

I'm viewing the article in Firefox as a sighted person and it's displaying as gibberish to me.

Even if it did work, if it's doing what I'm guessing it's trying to do it would be thoroughly unpleasant to read and I would leave the site immediately.

But agreed, it is not okay to intentionally sacrifice accessibility.

yes, it's an article about scrambling text using fonts in order to prevent LLM scraping.

It's a cute idea. I'd be more at ease if the author had described it as a neat toy that shouldn't be seriously used for anything, but they don't seem to be thinking that way, only describing the accessibility problems as a "drawback".

From the article "This works as long as scrapers don't use OCR for handling edge cases like this, but I don't think it would be feasible."

Lol, the article says it breaks copy and paste. I took a screen print, opened the image in macOS Preview, and selected the text.

A tech workaround for what this tech is doing is easy. Firefox and Acrobat both already do OCR. A scraper could do so as well.

hm-m, that sounds familiar… ah. Yeah. I've heard of this trick before.

Russia's Central Election Commission (ЦИК) pulled the same stunt with its website during the election, when that website published intermediate results, under the pretense of "preventing the system from crashing from unprecedented attacks"; I had to look up the date when exactly, seems to be 2021, which feels about right:
forbes.ru/tekhnologii/440479-p…

How was this supposed prevent any attacks and what *kinds* of attacks even is beyond me to this day. LLMs did not exist back then. Or in the very least weren't anywhere near as capable.

Памфилова объяснила кодирование результатов выборов на сайте ЦИК

Центризбирком закодировал результаты выборов на своем сайте, чтобы «система не вылетела» из-за обрушившихся на нее «беспрецедентных атак», заявила Элла Памфилова. Ранее эксперты обратили внимание, что копировать данные с сайта ЦИК стало невозможно, п

^{Тимур Батыров (Forbes.ru)}

In #Safari on an #iPhone going to the link I see... something that resembles a blog, but tapping the "reader view" I get the following, proving the point of the headline. If we can't read it, an #AI #Bot can't scrape it.

<begin text>
# /// script # requires-python = ">=3.12" # dependencies = [ # "bs4", # "fonttools", # ] # /// import random import string from typing import Dict from bs4 import BeautifulSoup from fontTools.ttLib import TTFont def scramble_font(seed: int = 1234) -> Dict[str, str]: random.seed(seed) font = TTFont("src/fonts/Mulish-Regular.ttf") # Pick a Unicode cmap (Windows BMP preferred) cmap_table = None for table in font["cmap"].tables: if table.isUnicode() and table.platformID == 3: break cmap_table = table cmap = cmap_table.cmap # Filter codepoints for a-z and A-Z codepoints = [cp for cp in cmap.keys() if chr(cp) in string.ascii_letters] glyphs = [cmap[cp] for cp in codepoints] shuffled_glyphs = glyphs[:] random.shuffle(shuffled_glyphs) # Create new mapping scrambled_cmap = dict(zip(codepoints, shuffled_glyphs, strict=True)) cmap_table.cmap = scrambled_cmap translation_mapping = {} for original_cp, original_glyph in zip(codepoints, glyphs, strict=True): for new_cp, new_glyph in scrambled_cmap.items(): if new_glyph == original_glyph: translation_mapping[chr(original_cp)] = chr(new_cp) break font.save("src/fonts/Mulish-Regular-scrambled.ttf") return translation_mapping def scramble_html( input: str, translation_mapping: Dict[str, str], ) -> str: def apply_cipher(text): repl = "".join(translation_mapping.get(c, c) for c in text) return repl # Read HTML file soup = BeautifulSoup(input, "html.parser") # Find all main elements main_elements = soup.find_all("main") skip_tags = {"code", "h1", "h2"} # Apply cipher only to text within main for main in main_elements: for elem in main.find_all(string=True): if elem.parent.name not in skip_tags: elem.replace_with(apply_cipher(elem)) return str(soup)
<end text>