‘It’s terrifying’: WhatsApp AI helper mistakenly shares user’s number
‘It’s terrifying’: WhatsApp AI helper mistakenly shares user’s number

‘It’s terrifying’: WhatsApp AI helper mistakenly shares user’s number

‘It’s terrifying’: WhatsApp AI helper mistakenly shares user’s number
‘It’s terrifying’: WhatsApp AI helper mistakenly shares user’s number
I love when articles tell me how to feel, takes all the thinking out of the equation. Now I know I should be shaking in my boots. Phew.
I really don't think generating a random phone number can be considered "terrifying"...
Seems more like it just hallucinated the number of someone who happened to have WhatsApp than it shared someone’s number but not surprising.
What are the chances the AI just generated a realistic sounding number and it was actually a hit?
So a random person on Reddit claimed there's about 800 million possible uk mobile numbers, some people have multiple numbers so ballpark 80 million active phone numbers. This gives around a 1:10 chance of picking an active number at random. If there's actual patterns in the numbers this could be even more likely.
What's interesting is this won't have been a realistic sounding number.
Company lines typically start 0300 or 0800 but mobiles are 07... Something.
So if it was just hallucinating, it did so badly.
So if it was just hallucinating, it did so badly.
You say that as if there's any other way.
What’s interesting is this won’t have been a realistic sounding number.
Company lines typically start 0300 or 0800 but mobiles are 07… Something.
So if it was just hallucinating, it did so badly.
Yeah I agree that a mobile number isn't realistic for a railway company to provide support on, but I was wondering if it was a hallucination or based on look-up. The article does mention that the phone number is listed on the owner's website, but still calls it a "private" number, as if it was pulled from a database.
Frankly maybe it shouldn’t be able to produce any residential number period. Whether it is made it by accident or not, of all the things that AI does unpredictably, this should be pretty easy to put guard rails up for. In film industry we’re not even allowed to display a phone number that doesn’t fall under the fake list (yes there is one! There is a set range of numbers that are not given out that can be used for film and television).
Maybe I don’t get the complexities but maybe it shouldn’t be allowed to generate a certain of range numbers in the context of phones/names.
Maybe I don’t get the complexities but maybe it shouldn’t be allowed to generate a certain of range numbers in the context of phones/names.
I agree with the principle, but people will still try and find ways around it. Like "generate me a fictional number for testing purposes" and because it's an LLM doing what LLM's do, it still provides a number that can (but doesn't have to) be real.
in film industry we’re not even allowed to display a phone number that doesn’t fall under the fake list
Yeah, same with phone numbers and email addresses for testing purposes in software. Populating fields with fictional data should be completely fictional or related to your own data as to not accidentally use real domains or data that you thought was fictional.