> because apparently what we have to offer is worth spending a ton of cycles to calculate the Anubis challenge.
I noticed the other day that lists.ffmpeg.org had moved to Anubis difficulty level 6, which takes ~180sec for my iPhone 17 to solve at ~100KH/s, making the site unusable. So I spent ~10 minutes vibe coding a safari extension with a native bridge to an optimized C kernel using ARM SHA256H* instructions that can do 200+ MH/s on the same device. This solves Anubis difficulty level 6 in a handful of milliseconds.
Given the numbers and capabilities involved (a single $5K ASIC miner yields 200TH/s, a million times more hash rate than my optimized kernel running on an iPhone), I don’t see how proof of work could possibly be a sustainable strategy to keep bots out without ruining human user experience. It’s an arms race that can’t be won.
Edit: I encourage you to try this yourself. Here's a sample prompt that ought to one-shot the task:
> Build an iOS Safari Web Extension that accelerates Anubis proof-of-work using a native C ARM64 SHA-256 kernel. Precompute the invariant 128-byte challenge prefix, search fixed-width decimal nonces with ARM SHA-2 intrinsics and two worker threads, and target difficulty-6 solves under one second. Relay challenges from a Safari content script through the background service worker to native code, then submit the valid nonce/hash through Anubis’s normal pass-challenge endpoint. Include a deterministic benchmark app, correctness tests against CryptoKit, bounded execution, and fallback to Anubis’s stock solver.For reference, the challenge on lists.ffmpeg.org takes 8 seconds on Firefox on my three year old laptop CPU that has worse benchmark scores than the iPhone 17 (tbf, the laptop also cost less than an iPhone 17). 8 seconds doesn't run against thermal limitations, so I really don't see why Safari on a modern iPhone should be so slow at this
> I really don't see why Safari on a modern iPhone should be so slow at thisThey're not trying to engage in an arms race, they're trying to channel a racing river into its natural course.
Developers and compute are interchangeable now.
I think the path forward is that websites offer one path for humans, and another for scrapers. But the huge catch is the path for scrapers must be _genuinely_ and _reliably_ the more economical and scalable path (either through something like PoW arms races, or through fear of litigation). Otherwise they will continue to ignore instructions and intrude on the human path.
(Also, blocking a whole country is likely not what you do, but you probably know that).
> so I really don't see why Safari on a modern iPhone should be so slow at this
FWIW, my iPhone 12 Mini also does ~110KH/s with Anubis on lists.ffmpeg.org, so seems fairly likely that Safari somehow here isn't working as expected.
Depending on the configuration, Anubis will supply a token after the challenge that bypasses the challenge for a time.
So any scraper that retains basic cookies will be able to bypass the challenge for a number of page views.
A user who needs to load a single page and a bot that wants to scrape a number of pages may pay the same cost.
The amortized per-view cost is highest for the real user.
Yes this needs to be tailored for Anubis, but that was the point: the moment it will become popular enough, scrapers will have no problems to bypass it.
The GPU cost to pretrain on that page once will dwarf by 2-3 OOMs the CPU cost to compute Anubis, scrape and post process it. And you’re not going to just train on it once!
At best you’re creating a speed bump for wannabe players scraping with no real plan. The folks training models people use just do not care.
Exactly, and that's even before accounting for more sophisticated PoW algorithms such as the one used by XMR that almost entirely eliminate any benefit from offloading to an accelerator.
Lots of work has been done to do just that by monero’s proof of work algorithm. They achieved ASIC and GPU resistance by making you generate entire programs (with quite a bit of RAM usage) and finding one that produces a desired output. Then you submit the seed used for the program (a series of programs actually but let’s not get into the details)
Anubis wants "slow for bots but fast for humans" which is not possible if somebody has a farm of devices that costs them basically zero.
That's not true. The goal is to raise the cost for bulk scrapers to the extent possible while minimizing any inconvenience to regular users. It's not about being slow for bots but rather reducing the asymmetry of the problem (from the perspective of the server) by imposing a (minor) cost on anyone requesting a page.
Combine with heuristics like in the article, so you can crank up the difficulty only for likely bot traffic, and you've got a bot supported site.
Opportunity costs are real.
I might try to tackle it again when I can give it some more attention.
the only feasible CAPTCHA at this moment is probably hCaptcha. they began to offer animated challenges. it's probably the best you can do, using an LLM to solve them would be a bit expensive for now.
At this point isn't it basically auth-only?
Rant: (genuinely wondering too, and RFC, request for conversation) at this point don't we have Google, etc. basically doing Real World ID Verification, but without an open protocol backing, using it to corral users into their ecosystem and gather data, and leaving us without some open and distributed alternative like the (fairly well working it seems to me) Public Key infrastructure? (PKI)
I know Sam Altman (with Proof of Human or w/e) and others have suggested and also have their own alternative (which I think is basically what Google already does), but I think this should simply be built based on an open protocol and then hopefully not-too-large (and with a healthy competition) local companies could do it for you.
All in the name of safety, of course.
Sure, the cheap lock I added to the outside shed's door-latch is terrible and can be shimmed open in two seconds with a bit of flat metal... but the important thing right now is that it keeps the raccoons out. If there's a new threat to the gardening tools or the bikes, then I have a different problem and I'll consider something else.
You can see the implementation here: https://gitlab.com/cryptsetup/cryptsetup/-/blob/main/lib/cry...
Or… the site will serve a random seed and the device must compute 4gb of pseudo-random data, then supply a value at a random server-demanded offset.
You yourself describe the solution.
Extensions or browser integrations of the Anubis task.
That's exactly not what happens with scrapers. Scrapers make millions of requests, yes. But each one of those requests produces an average (token-denominated) reward V'. There's no asymmetry at all: you're just adding a fixed cost to each reward, for both users and abusers.
That's before you even get to questions like "does the system actually work, or does a single solve get practically unbounded rewards".
There is: users solve the PoW once, scrappers solve it every time they rotate IP addresses. Some of them change their IP address on every request, so they keep solving the PoW over and over.
Which: I get! If it's your site, perform all the anger you want.
Anubis doesn't really help against this.
Even if you had to solve a L6 challenge for every request it's faster than the total RTT time of most servers. In other words not a meaningful barrier. And L6 is already a level which severely interferes with human usage of a site.
a $5000 ASIC-based device can hash SHA256 at 200TH/s or more.
If the source IP was an issue, you could do it other ways: for example, make the cookie rotate on every access, and insist there is a single stream of accesses.
You say “they” as if all scrapers are a monolithic group with the same constraints and goals. Part of the problem is the massive diversity.
And yet many report it works, at least for now, and the excess load due to scraping activity falls precipitously when Anubis or similar solutions are used. Maybe once this sort of challenge is used almost everywhere we'll see concerted attempts to get around it, but for now it is easier for them to just move on to another target.
> a $5000 ASIC-based device can hash SHA256 at 200TH/s or more
Peanuts for the big players, but many (almost all?) running smaller scale scraping operations are going to find $5000 rather prohibitive, and they are unlikely to be able to integrate it as they are probably running a “stock” scraper that they didn't write themselves.
Not sure why anyone would characterize these scrapers im general as all being fly-by-night operations that don’t have two cents to scrape together.
If they are so stupid an idea that they could never work, why do they seem to having the desired effect?
I wonder if AI might be able to come up with new PoW algorithms in a nightly CI job so every day is a different puzzle...
My earlier idea was to imagine that each day Anubis picks an entirely different problem-class. Ex: one day it is Sha256, the next it is prime factorization, the next it is twin-prime-finding, the next it is cracking elliptic curves, the next it is some kind of sorting / information theory problem...
All with the goal of adapting constantly so that scrapers have a harder time optimizing for the PoW problem (i.e. with Sha256 ASICs)
But it would just make Anubis suck a little bit less
Anubis "protection" model is wrong/broken on several levels. I suppose it does not even have ability to bypass hashing screen when server load is low.
Anubis is not a load optimizer. It's sole purpose is to deny the bots you don't want on your site.
It's on the same category as Iocaine and Go-Away, even if what they do is different in some cases.
This is fundamentally impossible.
In the end only pay per crawl will work. Like x402 but maybe something similar.
In other words: not only client has to spend some kind of resource for access, but it has to be a transaction, server must get something from serving the request.
No, this very article will tell you that the purpose of Anubis is to reduce load on the site. The purpose of denying bot requests is that you're strained by the load they impose.
Conversely, bots that pose no problems to you aren't "the bots you don't want on your site".
When the reward to abuse crosses a certain threshold, a determined party will come along and completely dominate it.
Anubis is just "solve hashes to get access to a valuable resources", and if the resource is at least valuable to whoever you're trying to deter as it is to legitimate users, you're not deterring anyone (and by extension, if it's more valuable to them, you're deterring your legitimate users instead).
Instead of the proof-of-work system of Anubis, I've gone down the iocaine route but implemented it in my application itself, as it's built in Elixir and causing problems for scrapers is really fun when it takes almost no server resources.
Currently I trick bad scrapers into a fake infinite black hole path with the promise of tasty data, then serve images to them one byte at a time over 15 minutes (after sending the header quickly), bloat the responses to cost them tokens, and randomly return AI generated images of sexy toasters. I have an admin dashboard with a little leaderboard for which ones get the most stuffed, and it keeps my heart warm on these wet autumn evenings.
You can't tease us like that without showing screenshots.
https://ibb.co/mCgXPZPg https://ibb.co/qFxVRQGW https://ibb.co/9m6xV1xg
I only just added the token wasting measure so I'm really curious to see what happens if another bot loads over 400 pages.
No hot toaster pics though :( on second thought perhaps better in our imaginations
I’ve also toyed with this approach and have a similar labyrinth on my website, but it just serves back huge amounts of meaningless text one word at a time.
The only punishment would be unique text that trains their models to be degenerate. And even then you'd probably have to serve across many domains.
The altered versions would be the page or paragraphs of it, with sentences and words swapped arbitrarily so that training on the data adds as many iffy token sequences as correct ones.
It'll just punish the bad actors running the scrapers. As the original poster mentions that they are using TVs as proxies to get residential IPs, one really can't think of these bots as criminal enterprises.
Sadly, if the bad actors has two cents for brain, they'll limit how much importance each domain name can have on training data. To mitigate impact of bad data like this.
(Note I'd suggest only linking to them from robot.txt as pages to not be indexed, that way no human or well behaved not ever will see them, which is kind of the point).
Exactly. I can't hope to keep them all at bay, but I can at least have the petty little victory of making their visit less convenient than it might otherwise be.
> if the bad actors has two cents for brain
I suspect that a majority of them are little better than the script kiddies of yore, running tools with minimal understanding of what is actually going on.
> I'd suggest only linking to them from robot.txt as pages to not be indexed
Agreed. Blocking all bots from all pages, well those that bother to listen to robots.txt. All bots because pretty much all of them are scraping for AI and similar these days, even googlebot. If I want people to see my stuff they'll get a link, and maybe they'll pass it on further, but all indexers/trainers can get stuffed. I'll likely make an exception for archive.org and similar.
Elixir's really fun to do this in because the BEAM will let you have hundreds of thousands of processes sat around doing nothing quite happily, so slow IO and such is something it's weirdly well-suited to.
This is on my photography site so I'm less fussed about them harvesting my writing, and more about causing problems for image harvesters. I'm sure they'll get some stuff anyway but at least one bot got stuck in a 400 page hellhole earlier today so I'm overall very pleased :D
I think the post underestimates just how little thought and effort is put into these bots. I also run a cgit instance with far less interesting projects, and am not spared from the deluge of HTTP requests.
The explanation I could come up with is that they try to crawl all links regardless of how much sense it makes or how much load it causes. cgit being cgit, this means billions of links for all combinations of parameters and hashes. That, or it's a deliberate DDoS attack.
https://news.ycombinator.com/item?id=44962529
It never really cohered as a solution. High-powered scrapers are better equipped to handle proof-of-work challenges than end users. Proof of work makes sense for a password hash, where any one guess at a password provides zero marginal utility. But every request from a scraper is productive to the scraper.
However, the refrain from naysayers isn't that there's too much enthusiasm. It's that anubis doesn't make sense as a theoretical barrier against bots. Who cares as long as it works empirically, even briefly? Are you suggesting people let their sites continue to be degraded or unusable until they find a better solution, even though PoW challenges can improve the situation today?
Not everyone has the luxury of writing their own web app designed to handle massive bot traffic at minimal cost, gutting functionality to meet that requirement if necessary.
I noticed a couple of weeks ago that 1-2 orders of magnitude more bots were getting past anubis challenges. However, it would be incorrect to conclude that it's doing nothing and now worthless aside from heating up visitors' cpus. If I turn it off, I see >2x bots.
The next salvo in the war of attrition will probably be when anubis switches to wasm hashx, which the author suggested, in previous HN threads, is on the roadmap.
If bots quickly adapt (Claude vibe-coded JIT wasm-to-native-transpiler for hashx challenge algorithms?), that may be the end of anubis and efforts like it. If they don't, I expect more HN threads with people complaining that such compiler efforts are possible and therefore PoW-gated sites are doing it wrong... even if those sites stay usable.
I'm just saying, it's bad computer science.
A better, perhaps the best, argument is that a worse bot mix might appear tomorrow, so anubis should be used as a runway for a better solution, and not as a solution in itself. If only everyone with a website had the desire (or AI tokens) to implement their own better solution. In some cases even desire or a lot of tokens isn't enough. I mean, you think kernel.org folks haven't put some serious thought into it? And yet, as HN argues about it, there's kernel.org running anubis at difficulty 5.
It shows they really don't care and they are just rage-baiting readers in their weekly grumpiness without any suggestions.
Unless they come up with said "solution", they can completely ignored in their complaints.
So could you give a short elaboration on what was this "better computer science" solution that you came up with in your discussion?
- It's the SOTA companies doing it?
- Scrapers are doing it for training data?
Those are two assumptions I see in posts and threads around Anubis, that are taken at faith, and never once substantiated.
[0] https://www.ft.com/content/07611b74-3d69-4579-9089-f2fc2af61...
There are more than just the American top dogs (OAI, Anthropic, SpaceX, Facebook)... especially the Chinese government with all its infinite cash resources and next to zero ethical constraints.
I don't trust the US top dogs at all, but I think the fear of discovery alone would lead them to not use "residential proxy" services. Non-US/EU entities however... who cares?
The scraper problem now is fleets of residential proxy devices - often things like smart TVs, phones, and browsers with some “proxy SDK” installed as part of an app’s monetization scheme. They make a couple of requests to a site - just enough to fly under the radar - and move on to a different site.
If each new site they hit forces them to solve a proof-of-work, that’s a meaningful dent in their scraping performance. Many of these boxes may not even have the spare CPU power to efficiently solve so many proofs of work - and anything that makes an owner notice their device is running slow is something that could meaningfully impede adoption of these SDKs, or force the operators to choose between minimizing performance impact or scraping more sites.
I just checked, and these many months later after installing Anubis, I'm getting about 14K requests per day. It's unreasonably effective at cutting crawler abuse. It should not be able to reduce my traffic by 98%, and yet in practice that's exactly what it's doing.
I think that the most abusive scrapers are operated by people who absolutely suck at programming. If they didn't, they wouldn't be crawling every permutation of "show me the git blame for every file in this repo for every commit in its history". Were they actually competent, they'd be investing in smarts such that they weren't crawling a web frontend to get that information.
See also how captchas continued being effective for years despite services like anti-captcha offering to solve them for you for a fifth of a cent each by farming the work out to India. It took advances in AI that made it viable to reliably solve them on-device to bring the end of the captcha
I don't think that's generally how these networks use residential exit proxies. There are at least a dozen well-developed frameworks out there for decoupling the crawler from the network exit point.
Most res proxy exits are just slinging bytes for clients using SOCKS, or another tunneling protocol.
If nothing else, a modern scraper will want better control over their TLS fingerprints, and you can't get that if you're depending on the on-device TLS libraries alone.
All they need to do is forward traffic and launder its IP origin. They don't even need to be able to (and would actually be well advised not to) decrypt TLS streams.
Until they discover this neat trick [1] and solve challenges orders of magnitudes more efficiently than legitimate users.
The game theory of Anubis is not sound. It makes fundamentally less sense than Captchas, and even those have been on the way out for a while.
Evidently, it makes _enough_ of a difference for the people running it. Criticising it is just letting the perfect be the enemy of the good.
The alternative to Anubis isn’t “something better that isn’t being picked”, it’s “completely unprotected”, and the people criticising it for being imperfect could helpfully move the conversation forward by actively suggesting a concrete alternative.
> meaningful dent
It also has an “online users” counter, which attempted to count real user sessions of unauthenticated user which still maintained a session, which lets them comment, or modify certain filter and display options. It never counted the Google bot.
Over the last few years this counter went from 100-200 users online to thousands. I have been very hands-off with it for many years, doing minor upgrades and backups. However, the site also has gotten quite slow, these sessions were obviously impacting it. So, I finally investigated these crawlers, and yes, it turns out it’s an insane amount of traffic that is entirely artificial, the site has just a handful of real users, and thousands of these crawling sessions that actually try to do everything they can, click every button. It doesn’t help that sort and search were implemented using GET links.
I fixed the counter to exclude the crawlers, but I have a bit of a dilemma. I don’t want to stop the bots from updating their knowledge based on all the content.
The best solution I could find is the new CloudFlare feature where they might charge the crawlers for every request, or otherwise block them. I think that’s a fantastic idea for the internet, at large. I signed up for the beta access, but haven’t heard from them again. I do think it’s unfortunate that this requires CloudFlare and the middleman.
Overall, it seems like the LLM are really straining the internet economy, the openness of it. Email spam used to be the worst, but the organized trillionaire labs sucking up the entire internet is going to break something if we don’t preempt them better.
It’s too bad the copyright and public internet systems are not acting quick enough. And I think there is no reason to act like this race really has to be at such a breakneck speed.
I would promote this idea to all of my competitors. Nah mate, you don't have to ask her out right now. You can wait until next week ;)
Anyway, silver linings, looks like we're finally going to get widely adopted infra for microtransactions.
https://web.archive.org/web/20030202042510/http://www.openp2...
Are the AI companies worried about anything but their valuations?
I don’t care about their valuations, but I do care about the risks that they are creating for the economy, society, and the technological advancement, at large.
Micro transactions [in this case] are a great idea, these crawlers need to be taxed and made to pay for the unaccounted external costs. Furthermore, we need to fund any risk mitigations, security coordination, and other safety work. If these things are not imposed on the market, the most ruthless player will create the biggest mess they can get away with.
I don’t see a better way to do this, right now, than to allow Strips, or CF, to create some way to charge the crawlers. Imagine I have some really expensive content, I can put it behind the paywall, or subscription model, but that’s very specific to the human consumer. What if I do want to search engine or content aggregator to consume some of the content? It would be quite hard for me to sign an exclusive agreement with OpenAI, like what Reddit or NYT are doing. It might be quite beneficial to create some infrastructure to let bots automatically to pay for access. It might even be crypto based, I don’t think it really matters. It just has to be easy for me to integrate it, and the bots need to find it easy to use.
If we all start doing this, I think this could be of benefit to all content creators, such as musicians, movie makers, game developers, and even independent journalists.
We used to pay for software and things, but what has happened? Walled gardens and production monopolies are taking over. This has largely reduced the quality of all the content we consume.
Anyway, what’s wrong with making 2 cents? :)
Why not? At this point, any attempt to try to meet the bots halfway is doomed.
Not as in “1 out of 1000 legit requests get blocked” but “you’re going to block 100% of some person’s requests.”
More over, I actually pay for three different AI subscriptions, and use them a lot. I also think it’s fair that I pay for their services. I just want them to also pay for what they use.
But I’m seeing requests coming in from random IP addresses from all over the world using randomised user-agents. But somehow these are trying URLs in sequential order. A while ago this would be classed as a DDoS cyber attack.
I feel like it's the same issue that public resources are facing.
Step 1: public resource is freely available
Step 2: asshole shows up, drains the resource completely just because they can
Step 3: either resource is taken away, or it has massive security behind it
We don't even need a new law. What they're doing is already a felony, and there's a process already in place to go to a court and request unmasking of 10000 IP addresses.
they have one of the worst bot detection in the industry. but because everyone uses them, their huge false positive numbers won't show up anywhere.
And Cloudflare is no exception, the "hello world" of bots is Cloudflare bypass.
It makes sense doesn't it? If you own all the pipes, you can effectively block bad actors. But in reality what they do is to just block everyone who blocks advertising tracking and profiling tech (such as canvas fingerprint etc)... and because all vendors use cf, they all block those same clients, and they think there's some winner takes all magic happening with finely tuned blocklists. Don't fall for the marketing!
Try it yourself. Block those fingerprinting tech and try to access their own user forum. you get blocked. Now allow fingerprinting (even if emulated in a headless chrome!) and access by the most toxic vpn you can think off, access granted.
All cloudflare does is act like a gatekeep for monetizable-by-creepy-advertising-networks traffic, not bots. If the bots pay the small headless chrome tax, you get bombarded with their requests just fine.
So you can definitely block by IP, by fingerprint, both or neither
And the bot progression from "alter user agent" to "change IP addresses" to providers having to ban whole subnets, whole ASNs, and realizing "proxy SDK monetization" is a thing mirrors threat actor progression from the time before LLMs.
Fork Anubis. Slightly modify the hash function it computes. Deploy. Do not try to make your fork widely adopted. Do not even publish it.
You've just defeated ASICs and any craweler that's special-cased Anubis (currently all of them). If enough people do this, the only recourse they will have is either genuinely executing served js code like a real user or building some unholy pipeline that uses ai agents to compile it to a GPU kernel for every host.
A more robust solution would be to keep a few patches handy with different versions of the algorithm, and rotate which one is in use. This would keep the crawlers on their toes if they wise up to the changed algorithm. One could even imagine automatically rotating witch algorithm to use on a weekly basis.
However day after day it just gets blasted with requests for deep pages. I was quite alarmed when I saw a 100x increase in the daily active user numbers which relied on session length, only to realize they were all bots.
Naively I too initially resorted to blocking user agents(Meta is thankfully nice enough to identify themselves, not nice enough to stop blasting 50k requests a day however), IP ranges from cloud providers and various browser fingerprints that I found connected to suspicious traffic.
However the battle seems unwinnable at the moment, outside of gating all content behind auth which I don't want to do. We have around 500k user generated content pages and I want those to remain publicly available.
I would be happy to provide our data to any one of these scrapers and I even added a message asking them to contact us if they want access to our data whenever I return a 403 response, however nobody has reached out.
Another campaign that someone is constantly running is daily checks for 100s of possible secret/config paths in hopes of finding an exposed private variable, these i've just blocked even though they would return a 404.
I still haven't found a way to deal with rotating residential IPs however, and most likely never will.
My current approach is to just run a 24 hour scan of all requests with codex and update my next.js proxy with more IP ranges, browser fingerprints and anything else that won't affect a real person.
Has anyone managed to come up with a way to stop this onslaught of crawlers and scrapers?
The AI crawler traffic I’ve seen sends one request per ip and seemingly has an infinite pool of residential IPs. You can’t block the ASNs becuase you also block honest clients. IP blocks are the wrong solution.
And because I’m being negative I’ll also be constructive, IMHO the correct solution for fighting residential proxy crawlers is using RTT diffs this is one example https://github.com/Sakura-sx/Aroma