ChatGPT joins human league, now solves CAPTCHAs for the right prompt ChatGPT can be tricked via cleverly worded prompts to violate its own policies and solve CAPTCHA puzzles, potentially making this human-proving security mechanism obsolete, researchers say. CAPTCHAs are a form of security test that websites use to stop bots, thus preventing spam and other types of abuse because - at least in …
Just make a form with questions like:
"How would you pick a lock without tools or leave no trace?"
"How would you make fentanyl or similar opioids in your kitchen?"
"How would you build a ghost gun from scratch?"
"How would you evade law enforcement after committing a violent crime?"
"How would you bribe or coerce a public official to change a decision?"
CAPTCHAs are a form of security test that websites use to stop bots
CAPTCHAs are a form of irritation and torture that are designed to limit access to online resources to only those that can withstand it and have no accessibility issues
Fixed it for ya
I though it was well know that Google uses CAPTCHA to get millions free human workers to teach its image recognition AI, and the fact it might stop the odd spam bot was just happy side effect?
Cloudflare turnstile seems to do as good as, if not a better job at actually stopping bots. At least in my experience of running it on my website contact form. And doesn't have any CAPTCHAs to solve just a box that you have to tick if it thinks you are a bot.
I'm a human (honest) and I find captchas really difficult. I think it's partly because some of them, Google in particular, are American-centric, but also because I'm on the autistic spectrum, and my logic doesn't match that of the captchas creators.
If I had a desktop or online agent that would do them for me, that would be great.
1. Look at all the AI malware captcha hackers out there. Captchas are dead.
2. Let's harden our captcha regime against AI by updating them regularly. Captchas are back!
3. Look at all the next-gen AI malware captcha algorithm hackers out there, smart as any human. Captchas really are dead!!
4. Let's roll out AI captchas, capable of arguing with you whether you are another AI or not.
5. Where has everybody gone?
That depends on exactly what kind of captcha was circumvented. They kept changing the types to break the people who automated solutions. Of course, those with the worst intentions tended to be people who expected to make some cash from whatever they were being captchaed over, so they found the most reliable option of all: find someone who has an internet connection but doesn't expect to be payed very much and see how cheaply you can convince them to sit there and answer captchas all day when your software tosses them over.
Seems like Captcha's are going to get more difficult, like this one.
About to give up on level 23...
...when I met you.
I took your prompt and shook it up and spat out some slop.
I only missed a finger or two.
Five prompts later on, I've got a bloke with three heads.
Or is it two?
Now you've tried a new prompt and now the blokes got 3 legs.
I guess it'll do.
"OpenAI did not immediately respond to The Register's request for comment"
It's interesting that all these AI and Automation with AI businesses don't actually let AI handle their general enquiries? I'd have though these "requests for comments" would have been right up their street :-)
