No, Apple’s AI isn’t reading your Signal Messages. But why are we so willing to believe it is?
I’ve been seeing a post go around the internet, and even in some private group chats, and I think you’ve probably seen it too. It’s about Apple Intelligence (Apple’s new AI product that comes standard on iOS 18.1) reading your Signal messages. It’s not entirely true, but its instructions won’t hurt anything. I want to unpack what Apple’s AI does, what it doesn’t do, and why we are all so eager to believe it can do things it can not.
The TLDR
Your Signal messages are just as secure as they ever were, but maybe turn off Apple Intelligence for other reasons. And don’t type anything you don’t want read in court, even if it is encrypted.
Here is the post
It reads “Apple can read signal messages on new Iphones! - If you have a newer iPhone, 15 or 16, and you have downloaded iOS 18.1, make sure you either disable Apple Artificial Intelligence all together under settings, or go under Siri/AI and ensure it is not enabled for Signal. If you do not, it will scan your Signal messages and read the content. Go to Settings > Siri > scroll down to Apps > Signal > turn off “Learn from this app” and the other two setting.”
I want to come out the gate saying you aren’t dumb for believing this, and nothing bad is going to happen because you did. It’s designed to prey on your emotions. But those emotions are interesting. They say more about our culture than they say about you.
I’ll also say, I fell for it the first couple times around. I don’t have an iPhone, so my instinct was to say “thanks for sharing,” and move on. But the 3rd time I saw it, in a different font, on a different platform, its virality triggered a fact checker in my brain. People with more tech experience than me have fallen for it too. The screenshot above is from a very tech-savvy sub-Reddit. I just want to be very clear that I do not want anyone to feel bad about all this.
(A note on terminology. I am going to oscillate between calling Apple’s AI product by its actual name “Apple Intelligence” what it actually is “Apple’s AI” and an abbreviation I made up “AAI”. It’s all the same thing.)
Apple’s Privacy settings
So, what data can Apple Intelligence read? We will use Signal as an example for how Apple could get the data from within your Signal app.
- You would have to have Apple Intelligence on. I suggest you just turn this off and don’t use it, but maybe you need it for something. I’m not your mom.
- You would need to give AAI permission to look at that specific app. This setting defaults to “on”. That is what this post is teaching you to turn off.
- The app’s developers would have to approve the data sharing. This is something Signal wouldn’t do, and can’t do, because of the nature of Signal’s encryption. Signal is open source, so we know they’re not lying about their encryption or data sharing.
So no, Signal is not sharing your data with AAI. AAI can not read your messages. But let’s look at another example. Maybe you use Safari as a browser. Safari is owned by Apple and not open source. We don’t know for sure what the developers would pick for step #3, but since they work for Apple, we can guess they would turn that data nozzle “on”. So if you don’t want that data collected (for AAI specifically) you could have to turn it off with either #1 or #2.
What Apple AI can do
What Apple Intelligence can do is a more complicated question. And I’m not really very interested in reading a year’s worth of ad copy to find out. Frankly, I think you should just turn it off, but newer iOS versions may not come with that option. If you’re interested in the changing AI field, I recommend reading pivot-to-ai.com.
What Apple AI can’t do… Or why are we all so eager to believe this post
So, why were we all so eager to believe this post? I think it is a combination of things. I think it has to do with all our weird feelings about how the world is changing around us.
It has to do with our weird feelings about AI.
We know it’s bad. The writer’s strike was partially about AI, it is super bad for the environment, it steals work from artists and writers, and we’ve all seen the slop it has produced clog the internet. So it’s not surprising to us when we’re told another bad thing it’s doing. The most direct example is Microsoft’s AI defaulting to “on” on Windows 11, and taking unencrypted screenshots every 2 seconds. (That product could theoretically read your Signal messages, if you have Signal desktop, so maybe turn it off.) So, why wouldn’t Apple to the same?
It has to do with our weird feelings about Apple.
Apple specifically has been in decline for years in public opinion. (I have no source for that but vibes.) New Apple products that people actually like are few and far between. They’ve stopped advertising themselves as the ‘privacy-focused’ tech brand. So people are growing skeptical, consciously or unconsciously. Capitalism grinds everything towards same-ness, so a once unique brand is now just like everyone else. And that leaves a sour taste in the mouths of people who supported them on the way up.
It has to do with our weird feelings about tech generally.
Tech is really complicated, but it is a huge part of our lives. And it is intentionally opaque. Tech has made billions through telling consumers, “this is the future, just buy in. We are smarter than you.” And these people are NOT smarter than you. We all have the sinking feeling in our belly that this is not what’s best for us, even if we can’t put our finger on it. When a post comes along that says “just follow these steps, and you’ll be safe from big tech,” it is really tempting. It makes us feel safe and helpful, if only for a second. And that is a value, even if it doesn’t keep you more private. But we need to be careful with that impulse, because similar impulses are powering QAnon and other destructive conspiracy movements.
It has to do with our weird feelings about capitalism.
I think subconsciously we all think there is some cheat code out of this hell-scape we are in. Like, if we just turn on the right settings or know the right person, or apply for the right government program, we won’t have to contend with the worst part of capitalism. It’s small in us, but when it blows up, it looks like “can I speak to your manager?” or Sovereign Citizens or fantasies about making a commune in the woods. That instinct is genuine. I constantly think “this can not be the system we all agreed to live under.” But it is. We live in a system of constant exploitation in many tiny and giant ways. So it’s tempting to see one of the tiny ways we’re being exploited, and a way to turn it off, and to do just that.
Conclusion
So ya. I have a lot of feelings about this post, and it will continue to circulate. But I want to re-iterate: You are not dumb for falling for this.
Call to action
Keep working on your BS-meter. Challenge the things you are willing to believe without evidence. And keep chipping away at all exploitation in any way you can.
Sources
I’m not going to @ anyone who posted this, but I’ve seen similar posts in all of the social media apps I use, even the more tech-savvy ones. This particular screenshot is from Instagram.
Some instructions on how to turn on and off Apple Intelligence: https://lifehacker.com/tech/how-to-enable-and-turn-off-apple-intelligence-on-iphone
More information on Apple’s general approach to privacy around AAI: https://9to5mac.com/2024/10/11/apple-intelligence-privacy-features-heres-what-you-should-know/
Recommended AI skepticism: https://pivot-to-ai.com/
As a post-note, there was a bit of a scandal a year ago when it came out that surveillance agencies were reading push notifications (the notifications that come up on your lock screen before you enter a passcode). Signal was never truly at risk for this, but other apps may be leaking data via this hacking method. So maybe turn off message previews on your lock screen. But I saw a similar panic, and that’s part of what made me so curious about this post.
News report on push notification vulnerabilities: https://www.reuters.com/technology/cybersecurity/governments-spying-apple-google-users-through-push-notifications-us-senator-2023-12-06/
Meredith Whittaker (President of Signal) discussing push notification security in a thread on Mastodon: https://mastodon.world/@Mer__edith/111563865413484025