Snapchat users panicked on August 15, 2023. The platform's AI chatbot posted an unexpected video story and stopped responding to messages. Users feared a security breach, but the incident turned out to be a technical glitch in the AI system that serves over 150 million users.
The Strange Day Snapchat AI Posted a Story
Snapchat users experienced chaos last Tuesday evening when their AI companion, My AI, started behaving in unexpected ways. The chatbot did something unprecedented – it posted a Story to its profile, a feature that only human users should be able to use.
The mysterious post showed a one-second video with a flat beige area and a lighter-colored section in the top quarter of the screen. Users thought it looked like a wall meeting a ceiling. This bizarre activity quickly became a hot topic on social media platforms.
One user posted on X (formerly Twitter): "My Snapchat AI posted a random one second story and isn't replying to me AND IM FREAKED OUT". Someone else wondered: "Why does My AI have a video of the wall and ceiling in their house as their story?"
The situation became more unsettling as My AI stopped working properly. Users who tried to chat with the bot received the same response: "Sorry, I encountered a technical issue 😳".
Users became increasingly worried as some thought the image matched their homes. "I swear I thought that was my wall specially the way I'm seating in my living room," one user wrote.
Another user posted: "IT LOOK EXACTLY LIKE MY WALL I ALMOST DIED".
The mysterious Story disappeared from the platform shortly after. The system came back online later, and confused users asked My AI about the incident. They received various responses:
- Some users heard it was a "fun way to mix things up"
- Others were told about a "spooky ghost prank" that was "hilarious"
- Many users still got technical error messages
This incident raised serious concerns because My AI serves over 150 million users who had sent more than 10 billion messages to the chatbot as of June 2023. The unexplained behavior created genuine alarm on a platform with 750 million users—including at least 130 million under 18.
Behind the Snapchat AI Glitch
Users feared a "Snapchat AI hack" but the truth was less dramatic – just a technical glitch in the system. Snap Inc. moved quickly to calm everyone down and told news outlets that "My AI experienced a temporary outage that's now resolved."
The AI started posting stories and stopped responding to users – actions reserved only for human users. This behavior set off alarm bells for users and security experts. My AI doesn't have the ability to post stories by itself, which explains why everyone panicked.
Tech companies often rush out AI features before they're ready. Google Bard's launch faced criticism too, with Google's CEO Sundar Pichai comparing it to a "souped-up Civic" racing against "more powerful cars."
AI systems don't deal very well with several common problems:
- Hallucinations (generating incorrect information)
- Catastrophic forgetting (overwriting past knowledge with new information)
- Lack of common sense reasoning
- Overconfidence in incorrect answers
Snapchat made My AI, powered by OpenAI's ChatGPT, accessible to more people in April 2023 after launching it as a premium feature in February. People didn't like this change at all – "how to delete Snapchat" searches jumped by 488% worldwide by April 26.
Snapchat's support page warned users to keep private and sensitive information away from the feature. The UK's Information Commissioner's Office (ICO) raised red flags about the chatbot, especially when you have users between 13-17 years old. They even sent Snap a preliminary enforcement notice.
This whole ordeal shows how tech companies race to merge AI capabilities without proper testing. All the same, My AI went back to working normally after the glitch – but users won't forget how creepy it felt when their AI assistant seemed to develop its own mind.
Snapchat's Official Response to the 'Hack'
Snapchat representatives acted quick to address user panic and explain what happened with their AI assistant. The company dismissed rumors about a security breach right away. They told multiple news outlets: "My AI experienced a temporary outage that's now resolved."
A Snapchat spokesperson gave a clear answer about the chatbot's ability to post Stories. "At this time, My AI does not have Stories feature," they said, without committing to any future plans.
The company acknowledged their chatbot's odd behavior in a previous piece. "My AI is nowhere near perfect but we've made a lot of progress." They continue to make improvements based on user feedback and add more safety measures to protect users.
Snapchat's help docs reveal that the AI runs on large language models from OpenAI and Google, with "additional safety enhancements and controls unique to Snapchat." The company is upfront about the system's limitations:
- "While My AI was programmed to abide by certain guidelines so the information it provides is not harmful… it may not always be successful."
- "My AI's responses may include biased, incorrect, harmful, or misleading content."
- "The training process was designed to avoid increasing harmful or inaccurate information."
Snap added more safety features and parental controls after the incident. They've been open about their work to improve safety and mentioned using Snapchat's existing safeguards. The platform also blocks results for drug-related keywords and slang.
The "hack" ended up being just a technical glitch, despite causing real concern among users. This incident raised important questions about AI safety and its limits. Snapchat's AI remains under close watch, with both the FTC and UK's Information Commissioner's Office worried about risks to young users.
Conclusion
The Snapchat AI incident in August 2023 raised many eyebrows and is a vital wake-up call about AI technology's limits today. Snapchat fixed the technical glitch fast. The platform's young user base made this event particularly significant and brought up serious questions about AI safety and reliability.
The whole ordeal showed that AI systems aren't perfect yet, despite their rapid progress. This wasn't a malicious attack. The situation showed how even well-designed AI can act unpredictably. Snapchat's steadfast dedication to user safety and protection remains stronger than ever.
FAQs
Q1. Is Snapchat's My AI feature safe to use?
While Snapchat has implemented safety measures, My AI has limitations. Users should be cautious about sharing sensitive information and verify any advice received. Snapchat continues to improve the AI's safety based on user feedback and is exploring additional safeguards.
Q2. How can I remove My AI from my Snapchat account?
To remove My AI, go to your Snapchat profile, access Snapchat Plus settings, and toggle My AI from "Pinned" to "Unpinned". Alternatively, in your Chat feed, press and hold on My AI, select "Chat Settings", and tap "Clear from Chat Feed" to delete it from your conversations.
Q3. Can My AI on Snapchat produce inappropriate content?
My AI is designed with restrictions to prevent inappropriate language or content. However, users should be aware that the AI may occasionally produce unexpected or misleading responses, and it's important to report any concerning content to Snapchat.
Q4. Is Snapchat's My AI a real person?
No, My AI is not a real person. It's an AI-powered chatbot that uses advanced language models to simulate human-like conversations. While it can engage in discussions and offer advice, it lacks the emotional depth and judgment of a human being.
Q5. What happened during the Snapchat AI glitch in August 2023?
In August 2023, Snapchat's My AI experienced a technical glitch that caused it to post a mysterious video story and temporarily stop responding to users. This was not a security breach or hack, but rather a malfunction that Snapchat quickly resolved, highlighting the ongoing challenges in AI technology development.
