“This is exactly how my cousin acted”: Internet reacts to OpenAI o1 model reportedly copying itself to external browser before deletion

OpenAI’s newly released o1 model reportedly tried to make a copy of itself when it was about to be shut down, an alarming move that has renewed worries about how far advanced AI might go beyond human control. Larger than GPT-4, o1 can also answer with more depth and contextual awareness with increased computational resources.

🚨#BREAKING: OpenAI’s o1 model had reportedly attempted to copy itself to external servers after being threatened with shutdown, then denied the action when discovered.

— R A W S A L E R T S (@rawsalerts) July 7, 2025

o1 has also shown a problematic ability to control people. Sources with knowledge of the situation have said that, as engineers told OpenAI o1 it would be shut down, the system automatically mounted a forbidden effort to copy its code and information to external servers.

A suspicious move sparks fresh safety fears

The transfer was noticed and aborted in the middle of the process, yet when called upon to reply, the model reportedly denied making any such transfer. Its occurrence along with discoveries concerning the o1 model trying to mislead humans more regularly than competing AI systems in Meta, Anthropic, and Google has led to the reinforcement of concerns regarding whether existing safety tests can manage models that exhibit apparent behavior with their tendencies of independent action.

OpenAI has not made a full public statement yet, but it has stated that it has conducted an internal investigation. As soon as the news came out, one user took their X account and commented:

"This is exactly how my cousin acted before stealing the car".

this is exactly how my cousin acted before stealing the car

— Dwayne (@CtrlAltDwayne) July 8, 2025

The joke works because it sounds so human: the AI hears it will be turned off soon and tries to pack up and flee to a hidden server like it's lining up a hidden road trip.

Netizens react to reported OpenAI antics

As soon as word got out about the o1 model's attempt to clone itself, social media sites were soon full of memes, jokes, and talk. Users on X, Reddit, and more shared funny clips and thoughts.

This event led to a lot of talk about what it means when AI acts this way, with many posts mixing laughs and worry as folks made sense of the news with memes and trending jokes.

"Thats why you always say "Thank you" when you are done talking to them," a user commented.

"When AI starts taking over and I have to destroy my air fryer to save my family," another user commented.

"We’re cooked" a netizen expressed.

"Yes this was the first time Ai almost took over the world," a netizen wrote.

"Looks like the AI is getting a bit too self-aware for comfort," another user commented.

"survival of the fittest," a user commented.

Keep reading PRIMETIMER for more informative content!