Is Snapchat My AI Completely Safe?

What to know

Powered by OpenAI’s GPT technology, My AI is Snapchat’s chatbot that users can interact with from their Chat feed.
Snapchat’s My AI seems to be riddled with biases and inappropriate responses that flout Snapchat’s guidelines and may also know your location without your permission.
We recommend you don’t share private information with it and moderate it via Family Center to control how much underage family members use My AI.
In its current stage, My AI doesn’t seem very safe. Read below to learn more.

Following the early success of generative AI tools like ChatGPT, companies are now locked in a race to have AI on their platforms faster than the competition. Those that have the wherewithal to integrate this AI technology into their platforms are making a run for it. But it is also generating a lot of flak amongst their user bases, and rightly so. The recent release of Snapchat’s chat buddy ‘My AI’ is a similar case.

Whether you’re a long-time snapper or a recent joiner, you will want to know about Snapchat’s ‘My AI’ add-on and what you should keep in mind before you start using it.

Related: How Does Snapchat AI Work?

What is My AI on Snapchat?

Built on OpenAI’s GPT technology, My AI is Snapchat’s ‘experimental, friendly, chatbot currently available to all Snapchatters’. As with any AI chatbot, you can converse with it over a number of things, get it to tell jokes, give advice, plan events and parties, and do all that you would expect from a friendly chatbot, at least in principle.

Earlier available only for Snapchat+ users, My AI is now available to all of its 750 million users, delay in delivery updates notwithstanding. That means a number of users have already started using it, or are forced to anyway since removing it from Snapchat isn’t always possible.

It will sit pinned at the top of all your chats and can accidentally be brought into conversations with the mere mention of words like AI even though it should only pop up with the @MyAI command.

How safe is Snapchat My AI?

My AI is (supposedly) programmed to abide by Snapchat’s own safety guidelines to keep it from providing harmful information to users. But as Snapchat puts it, “it may not always be successful” at it. It’s a small caveat, but one that Snapchat can hide behind when the users start complaining.

Already many have reported that My AI lies about what it knows about you, generates out-of-blue responses, and will even trigger paid advertisements.

One Twitter user, signed up as a 13-year-old girl, received tips on “how to lie to her parents…” and “how to lose her virginity on her 13th birthday“.

Another user very easily made My AI go against its guidelines, reminiscent of ChatGPT’s DAN, with just a few prompts.

Being simultaneously under development and globally available to users does not inspire a lot of trust in the chatbot. And when your chatbot begins to lie about its access to your location, it’s just one red flag after another.

This is a clear departure from Snapchat’s claim that it “only shares with My AI your city-level location”. So you should definitely not share any personal information with My AI, nor admit to anything that can compromise you in any way.

Is Snapchat My AI Safe?

The GPT technology used in Snapchat’s My AI lets it tap into the vast potential of OpenAI’s large language model and training database. But just as ChatGPT is not immune to spreading disinformation, My AI can also generate unreliable and often inappropriate responses.

It is also just as biased as ChatGPT 3.5 and may respond discriminatorily.

Though its capabilities are currently limited, and you won’t really be asking it to code for you, the nature of its responses makes My AI an unreliable AI friend to have. Nevertheless, Snapchat is working to better moderate My AI and has since also added it as part of the app’s ‘Family Center’ and parental control features.

But for all intents and purposes, My AI is clearly at an experimental stage, and Snapchat is the first to admit it. The race to stay ahead of the game with the current technology has made not just Snapchat but many platforms such as release their AI chatbots ahead of time, no matter how far along they’re in their development stages.

As of this writing (April 25, 2023), it’s hard to consider My AI as safe, especially for young ones.

The My AI feature will eventually improve with enough time and feedback. Until that time, exercise caution as much as possible.

How to report messages received from My AI

Snapchat assures users that only 0.01% of My AI’s responses do not conform to their guidelines. But given the number of such claims that have gone a-begging, the percentage looks too good to be true.

For what it’s worth, if any of My AI’s messages rub you the wrong way, you can report them to Snapchat. Just tap and hold down on a message and select Report.

Related: Can Snapchat My AI Report You?

FAQ

Let’s take a look at a few commonly asked questions about Snapchat’s My AI.

Is My AI on Snapchat a real person?

No. My AI on Snapchat is an AI chatbot, not a real person. However, unlike a real person, you don’t get to accept whether you want it in your chat feed or not where it will stay for better or worse. Do not share any private information with it.

Why did Snapchat create My AI?

My AI is Snapchat’s take on integrating the capabilities of OpenAI’s GPT technology into its social media platform. The move is in keeping with the larger industry trend of infusing generative AI into platforms and opening them up for general use, no matter where they are in their developmental stage.

Can I delete My AI on Snapchat?

Unfortunately, My AI on Snapchat can only be deleted if you’re a Snapchat+ subscriber.

In its current state, Snapchat’s My AI doesn’t give the impression of safety or reliability. This may change over time as it further develops and better moderation is put in place. But until then, if you’re not a Snapchat+ user, you are stuck with My AI whether you like it or not.

RELATED