On Tuesday, OpenAI announced that it was shutting down Sora, its AI video creation. However, it appears that the app isn't the only project the company is canceling this month.
In October, OpenAI head Sam Altmanannounced plans to introduce an “erotica mode” to ChatGPT. The feature was an add-on to an otherwise unrelated announcement about age verification.
See also: OpenAI desktop “superapp”: Combines ChatGPT, Codex and Atlas
In December, OpenAI implemented age verification and announced that, as part of its effort to “treat adult users like adults,” it planned to enable even more, such as an erotica mode for verified adult users.
However, the “spicy” version of ChatGPT didn’t reach users in December, and on March 6, Alex Heath reported that the company was “freezing” “Adult Mode” in the AI chatbot. Instead, OpenAI said it would prioritize improvements to intelligence.
The Financial Times later reported that the pause is “indefinite,” citing challenges such as “sexual datasets and problems in removing illegal content.”
See also: Encyclopedia Britannica sues OpenAI for use of its content
OpenAI: Reassessing Priorities
Recent developments clearly show that OpenAI is reassessing its priorities in a period of intense pressure and increasing scrutiny. The shutdown of Sora and the “freezing” of the so-called Adult Mode are not just isolated decisions, but are part of a broader strategy of folding, aimed at concentrating resources on improving the basic intelligence of the models.
Despite initial announcements of more "open" capabilities for adult users, the reality of technical and ethical challenges—such as managing sensitive datasets and preventing illegal content—seems to dictate more conservative choices.
See also: ChatGPT makes it easier to understand mathematical concepts
At the same time, this stance highlights a crucial dilemma for the future of artificial intelligence: to what extent can freedom of use be extended without compromising security and reliability?
March proved to be a watershed month for OpenAI, with decisions that may have disappointed some in the public eye but potentially averted greater reputational and regulatory. In an environment where AI is evolving faster than regulatory frameworks, cautious slowing down may ultimately prove more strategic than unfettered innovation.
