The article is more of an intro piece for newcomers and doesn't discuss at all the state of the art or where the competition is--the high end of the market is pretty saturated already but the low end is wide open.
There is a bit of a spread in the market, and the specific detection techniques are ofc proprietary and dynamic. Until you have stewed on it quite a bit, it is reasonable to assume everything you can think of has a- been tried b- is either mainstream or doesn't work well c- what working well means is subtle.
Bots are adversarial and nasty ones play the field. Sources of truth are scarce and expensive to consult, and the costs of false positives are felt acutely by the users and the buyers, vs false negatives are more of a slow burn and a nagging suspicion.
If they can narrow down the possibilities to quadratic space then you lose.
Marketing indeed. He had me doubting for a while what magic they weren't sharing with the rest of us to avoid countermeasures being developed, but I know better now (working in infosec, seeing what these systems catch, don't catch, and bycatch)
I can tell you that as soon as you download Chrome and login to any Google account of yours, the captcha tests are suddenly and mysteriously gone.
Use firefox in full-lockdown mode, and you will be clicking fire hydrants and crosswalks for the next several hours.
My crazy conspiracy theory is that Google is just using captcha as an opportunity to force everyone out of privacy mode, further empowering the surveillance capitalism engines. The intent is not to be effective, but inconvenient.
Very few sites are broken by blocking Google's features, incidentally. Even Privacy Badger warns that blocking Google Tag Manager may break sites. It doesn't break anything important.
https://en.wikipedia.org/wiki/Goodhart%27s_law
BRB, changing the simulated latency in my bot.
Proof-of-work CAPTCHAs work well for making bots expensive to run at scale, but they still rely on accurate bot detection. Avoiding both false positives and negatives is crucial, yet all existing approaches are not reliable enough.
One comment re:
> While AI agents can theoretically simulate these patterns, the effort likely outweighs other alternatives.
For now. Behavioral and cognitive signals seem to work against the current generation of bots, but will likely also be defeated as AI tools become cheaper and more accessible. It's only a matter of time until attackers can train a model on real human input, and inference to be cheap enough. Or just for the benefit of using a bot on a specific target to outweigh the costs.
So I think we will need a different detection mechanism. Maybe something from the real world, some type of ID, or even micropayments. I'm not sure, but it's clear that bot detection is at the opposite, and currently losing, side of the AI race.
I think the most likely long-term solution is something like DIDs.
https://en.wikipedia.org/wiki/Decentralized_identifier
A small number of trusted authorities (e.g. governments) issue IDs. Users can identify themselves to third-parties without disclosing their real-world identity to the third-party and without disclosing their interaction with the third-party to the issuing body.
The key part of this is that the identity is persistent. A website might not know who you are, but they know when it’s you returning. So if you get banned, you can’t just register a new account to evade the ban. You’d need to do the equivalent of getting a new passport from your government.
Are you sure you read my comment fully?
the governments powerful enough to roll something like this out are not trusted authorities which will protect the privacy of their citizens. remember before the Snowden revelations when the NSA's director of national intelligence swore under oath that they did not collect "any type of data at all on millions of Americans"?
https://en.wikipedia.org/wiki/James_Clapper#Testimony_to_Con...
This concept has been studied already extensively, e.g [1] (in 2000!) by people like Rivest and Chaum, who have actual decade-old competence in that field.
[1] https://people.csail.mit.edu/rivest/pubs/pubs/LRSW99.pdf
https://www.nytimes.com/2006/02/05/technology/postage-is-due...
Are you sure? And how do you know?
There are a lot of CAPTCHA cracking services. Given the price, they are hardly sustainable even under developing country wage level. I believe they actually solve the easy ones automatically and humans are only involved for the harder ones.
The ultimate challenge is to replicate end-to-end natural human cognition, which is currently an unsolved and hard problem (and also not necessarily the main focus of AI researchers).
This is wrong, badly wrong.
CAPTCHA stood for “Completely Automated Public Turing test to tell Computers and Humans Apart”. And that’s how people are using such things: to tell computers and humans apart. But that’s not the right problem.
Spam and abuse can come from computers, or from humans.
Productive use can come from humans, or from computers.
Abuse prevention should not be about distinguishing computers and humans: it should be about the actual usage behaviour.
CAPTCHAs are fundamentally solving the wrong problem. Twenty years ago, they were a tolerable proxy for the right problem: imperfect, but generally good enough. But they have become a worse proxy over time.
Also, “human-friendly CAPTCHAs” are just flat-out impossible in the long term. As you identify, it’s only a “for now” thing. Once it’s a target, it ceases to be effective. And the range in humans is so broad that it’s generally distressingly easy to make a bot exceed the lower reaches of human performance.
> Proof-of-work CAPTCHAs work well for making bots expensive to run at scale, but they still rely on accurate bot detection. Avoiding both false positives and negatives is crucial, yet all existing approaches are not reliable enough.
Proof-of-work is even more obviously a temporary solution, security by obscurity: it relies upon symmetry in computation power, which is just wildly incorrect. And all of the implementations I know of have made the bone-headed decision to start with SHA-256 hashing, which amplifies this asymmetry to ludicrous degree (factors of tens of thousands with common hardware, to tens of millions with Bitcoin mining hardware). At that point, forget choosing different iteration counts based on bot detection, it doesn’t even matter.
—⁂—
The inconvenient truth is: there is no Final Ultimate Solution to the Spam Problem (FUSSP).
> Productive use can come from humans, or from computers.
I agree in principle, but the reality is that 37% of all internet traffic originates from bots[1]. The overwhelming majority of that traffic (89% according to Fastly) can be described as abusive. In turn, the abusive traffic from humans likely pales in comparison. It's vastly cheaper to setup bot farms than mechanical turk farms, and it's only getting cheaper.
Identifying the source of the traffic, while difficult, is a generalizable problem. Whereas tracking specific behavior will depend on each site, and will likely require custom implementation for each type of service. Or it requires invasive tracking of users throughout the duration of their session, as many fraud prevention systems do.
Both approaches can be deployed at the same time. A CAPTCHA is not meant to be the only security solution anyway, but as a first layer of defense that is generally simple to deploy and maintain.
That said, I concede that the sentence "[CAPTCHAs] are the only solution" is wrong. :)
> Proof-of-work is even more obviously a temporary solution, security by obscurity
I disagree, and don't see how it's security by obscurity. It's simply a method of increasing the access cost for abusive traffic. The more signals are gathered that identify the user as abusive, the higher the "price" they're required to pay to access the service. Whether the user is a suspected bot or not could just be one type of signal. Behavioral and cognitive signals as mentioned in TFA can be others. Yes, these methods aren't perfect, and can mistakenly penalize human users and be spoofed by bots, but it's the best we currently have. This is what I'd like to see improved.
Still, even with all their faults, I think PoW CAPTCHAs offer a much better UX than traditional CAPTCHAs ever did. Yes, telling humans apart from computers is getting more difficult, but it doesn't mean that the task is pointless.
[1]: https://learn.fastly.com/rs/025-XKO-469/images/Fastly-Threat...
“Expensive” depends on the value of what you do behind the captcha
There are human-solving captcha services that charge USD 1 for 1k captchas solved (0.1 cents per captcha)
So as long as you can charge more than what solving the captchas cost, you are good to go
Unfortunately, for a lot of tasks, humans are currently cheaper than AI
2. Send link to coworkers via Slack so they can spend five minutes doing the tasks.
3. Capture that data and create thousands of slight variations saved to db as profiles
4. Bypass bot protections.
There is nothing anyone can do to prevent bots.
No they don't, that's the point: you can serve everyone a PoW and don't have to discriminate and ban real people. This system you're enthusiastic about is what tries to do this "accurate bot detection" (scratch the first word)
https://github.com/TecharoHQ/anubis/blob/main/data/botPolici...
Distinguishing en mass seems like a waste to me. Deal with the actual problems like resource abuse.
I think part of the issue is that a lot of people are lying to themselves that they "love the public" when in reality they really don't and want nothing to do with them. They lack the introspection to untangle that though and express themselves with different technical solutions.
I work for Stytch and for us, that looks like:
1) make it easy to provide Connected Apps experiences, like OAuth-style consent screens "Do you want to grant MyAgent access to your Google Drive files?"
2) make it easy to detect all bots and shift them towards the happy path. For example, "Looks like you're scraping my website for AI training. If you want to see the content easily, just grab it all at /LLMs.txt instead."
As other comments mention, bot traffic is overwhelmingly malicious. Being able to cheaply distinguish bots and add friction makes your life as a defending team much easier.
Have you never once looked at the captcha and couldn't decide whether the 3 pixels of the motorcycle sticking out into the grid square meant that you should select that grid square too? Not once? As the tests become ever more sophisticated, more and more of you all will be locked out.
... break the Web.
ftfy
I have the opposite view. This already played out in the Minecraft community and it turns out ghost clients are effective in spoofing such behavioral signals and avoiding anticheat. Also I doubt you can get any meaningful signal from the couple of a seconds a user's ai agent is scrolling through a site.
Whether the person interacting with your website is human or not isn't relevant anymore. What matters is what they are doing; be they human, bot, or AI agent.
See also Elon demanding ad spend on his platform or like it’s literally just like when the Nazis invaded Poland. Anyone got some E? PLUR, bro but also fewer vacay days for you.
Empty economic activity driven by fiat decree of wealth hoarders suffering from post war and Cold War and leaded gas fume, lead water fueled paranoias and psychosis.
People made insane by memorization of illusory social obligations to history always run the world.
Like the case recently of builder.ai which had humans pretending to be ai.
Turing was a visionary - but even he could not imagine a time when humans pretend to be bots.
With a hobby wiki, eventual consistency is fine. I believe ghost bans and quarantine and some sort of invisible captcha would go a long way toward my goal, but it’s hard to find invisible captcha.
There was a research project long ago that used high resolution data from keyboards to determine who was typing. The idea was not to use the typing pattern as a password, but to flag suspicious activity. To have someone walk past that desk to see if Sally hurt her arm playing tennis this weekend of if Dave is fucking around on her computer while she’s in a meeting
That’s about the level I’m looking for. Assume everyone is a bot during a probationary period and put accounts into buckets of likely human, likely bot, and unknown.
What I’d have to work out though is temporary storage for candidate edits in a way they cannot fill up my database. A way to throttle them and throw some away if they hit a limit. Otherwise it’s still a DOS attack.
You always have to show people their own edits. It's a common form of proofreading. But what's added and how often does matter. Misinformation is one thing. External links are potentially something much worse. I used to think SO had it figured out as far as mutual policing, but that's not working so well now either.
I agree that better authentication methods for AI agents are needed. But right now bots and malicious agents are a real problem for anyone running sites with significant traffic. In the long run I don’t think human traffic will go to zero even if its relative proportion is reduced.
I want it. I don't want my message boards to be people's AI agents...
These tests here are easily bypassable, just adding a random delay somewhere during the action phases to mimic humans, and there's already tools for mimicking human mouse movements.
> what if the turing test already runs silently across every site you open. just passive gating based on scroll cadence, mouse entropy, input lag without captcha or prompt
>what if you already failed one today. maybe your browser fingerprint was too rare, maybe your keyboard rhythm matched a bot cluster from six months ago. so the UI throttled by 200ms. or the request just 403'd.
> what if the system doesn't need to prove you're a bot. it just needs a small enough doubt to skip serving you the real content.
> what if human is no longer biological but statistical. a moving average of behavior trained on telemetry from five metro cities. everyone outside that gets misclassified.
>what if you'll never know. timeline loads emptier than someone else with explicit rejection to the content
I use keyboard navigation on many pages. Using the firefox setting "search when you start typing", I don't have to hit ctrl+f to search on the page, I just type what I want to click on and press enter or ctrl+enter for a new browser tab, or press (shift+)tab to go to the nearest (previous/next) input field. When I open HN, it's muscle memory: ctrl+t (new tab) new enter (autocompletes to the domain) thr enter (go to threads page) anything new? type first few chars of username, shift+tab+tab enter to upvote. Done? Backspace to go back. View comments of a link? Type last char of a word in the link, space, and first char of next word, that's almost always unique on the page, then escape, type men, enter, to almost always activate the comment link. Or shift+tab enter instead to upvote. On the comments page, reading top-level comments is either searching for [ and then enter+f3 when I want to collapse the next one, space for page down... Don't have to take my hands off the home row
etc. on lots of website, also ones I've never visited before (it'll be slower and less habitual of course, but still: if there is text near to where I want to go, I'm typing it). I use the mouse as well, but I find it harder to use than the keys that are always in the same place, much easier to press
So will it tell me that my mouse movements don't look human enough or will I see a "Sorry, something went wrong" http 403 error and have no clue if it's tracking cookies, my IP address, that I don't use Google Chrome®, that I went through pages too fast, that I didn't come past the expected page (where a cookie gets set) but clicked on a search result directly, that I have a bank in country A but residence in country B, that I now did too many tries in figuring out which of these factors is blocking me.... I can give examples of websites where I got blocked in the last ~2 months for each of these. It's such a minefield. The only thing that always passes is proof-of-work CPU challenges, but I dread to think what poor/eco people with slow/old computers are facing. Will this "invisible" captcha (yeah, invisible until you get banned) at least tell me how I'm supposed to give my money to whatever service or webshop will use this?
It might work for a while, but that's a losing battle.
Hi this is incorrect. Different =/= dumber. The insight is that humans and computers have different constraints / algorithmic capabilities / objective functions / etc.
Maybe it was my fault to advertise my own solution in comments.
Such behavior however triggered bot detection. I might have behaved like a NPC. So currently a human can be identified as a bot, and banned on that premise. Crazy times.
Currently I feel I must act like a human.
BobbyTables2•5h ago