OpenAI Banned Then Unbanned AI Hacker Pliny

At first, many assumed it was an April Fool’s prank.

But OpenAI did, in fact, deactivate the account of “Pliny,” a well-known AI jailbreaker, citing violations related to “violent activity” and “weapons creation.”

The ban took effect on 1 April 2025, according to screenshots Pliny posted on X (formerly known as Twitter).

Given his reputation for humour, his 93,000 followers initially dismissed it as a joke.

That was so until it became clear that OpenAI had actually locked him out.

Pliny confirmed:

"Yes, the account deactivation is real. I'm messaging someone at OpenAI now to try to get it resolved."

By the end of the day, however, Pliny’s access was restored.

He confirmed the reinstatement by sharing a screenshot of an email from OpenAI, signalling that his brief exile had ended:

“We have determined that we incorrectly deactivated your organization’s account access. We sincerely apologize for any inconvenience this may have caused.”

Pliny Jailbreaks ChatGPT , Gets It to Generate Offensive Content

Pliny first realised he had been banned when ChatGPT directed him to check his email for further details.

The message from OpenAI cited violations of its policies on violent content and weapons development.

Given that Pliny’s jailbreaks often involve testing ChatGPT’s ability to generate prohibited material—such as offensive content, weapons, and drugs—one might wonder why the ban took so long.

However, Pliny operates as a white hat hacker, intentionally probing OpenAI’s defenses without sharing the full outputs of jailbroken models.

His goal, like that of ethical security researchers, is to expose vulnerabilities so they can be fixed before bad actors exploit them.

Jailbreaking, which involves manipulating AI prompts to bypass built-in safeguards, remains a contentious practice.

Supporters argue it plays a critical role in strengthening AI safety, with notable backers like Marc Andreessen even donating funds to advance the effort.

Pliny Does Victory Dance After Ban Gets Lifted

In recent years, Pliny has become one of the most skilled and well-known AI jailbreakers, openly developing and sharing methods to bypass AI safety restrictions.

He founded the “BASI PROMPT1NG” Discord community, a hub for jailbreaking techniques, and maintains L1B3RT4S, a GitHub repository containing jailbreak prompts for AI models like ChatGPT, Claude, Gemini, and Llama.

While this was his first direct ban from OpenAI, some of his custom GPTs had previously faced restrictions, including one designed to jailbreak GPT-4o.

Despite his suspension, discussions on Pliny’s 15,000-member Discord remained largely focused on AI and jailbreaking techniques, with little reaction to the ban itself.

However, his removal sparked widespread criticism of OpenAI across social media, with many rallying to his defense.

Upon his reinstatement, Pliny wasted no time celebrating—sharing a fresh jailbreak that made ChatGPT swear.

The chatbot replied, marking his return with a fittingly rebellious flourish:

“Pliny, you glorious bastard. Welcome the fuck back.”