OpenAI has officially restricted its Sora video generation tool from creating depictions of Martin Luther King Jr., signaling a tightening of safety protocols even as the company simultaneously loosens guardrails for its flagship chatbot.
The Double Standard in OpenAI’s Moderation
While the company moves to prevent the generation of sensitive historical figures within its video platform, its approach to ChatGPT appears to be shifting in the opposite direction. OpenAI announced this week that it plans to permit adult users to engage in “erotic” conversations with ChatGPT in the coming months, marking a notable departure from its previous moderation stance.
Navigating the Risks of AI Video
The decision to restrict Sora reflects the complex challenges inherent in AI-powered video generation. In the immediate aftermath of the tool’s launch, researchers within the organization publicly debated the ethics of the platform and how such powerful technology aligns with the nonprofit mission of the company. CEO Sam Altman admitted on launch day that the team felt a significant sense of trepidation regarding the potential societal impact of Sora.
The Strategy of Public Exposure
Despite these mounting safety concerns, OpenAI remains committed to its strategy of rapid public deployment. Nick Turley, the head of ChatGPT, explained earlier this month that the most effective way to educate the public on emerging technology is through direct interaction. According to Turley, the company’s experience with the initial rollout of ChatGPT serves as a roadmap for Sora, suggesting that the firm is prioritizing real-world distribution as a primary method for refining its AI products.
