Openai has announced changes to its content moderation policy. Currently, ChatGpt is allowing images of public figures, controversial symbols, and racial traits. Previously, the company was reluctant to generate such content, citing concerns about potential harm and controversy.
The revised policy revealed in this week’s blog post by Openai’s model behavior lead Joanne Jang, shows the key evolution of the company’s approach to content moderation. Chang explained that Openai is moving away from blanket rejection due to sensitive topics and adopting a more nuanced strategy focusing on minimizing real-world harm.
“We are moving from rejecting sensitive local blankets to a more accurate approach focusing on preventing real-world harm,” writes Chan. “The goal is to embrace humility. To recognize how much we don’t know and to place ourselves in an adaptation as we learn.”
The updated policy allows ChatGpt to create images of prominent figures, including former US presidents Donald Trump and Elon Musk, who were previously prohibited from the system. This new feature allows users to decide whether images can be used by AI generation through OpenAI.
Now, Openai allows for the generation of “hate symbols” such as sw, which have specific conditions that contain educational or neutral contexts that do not support extremist movements. Openai now accepts content applications that were previously completely prohibited.
The updated policy changes what the company counts as offensive content. AI previously rejected requests for physical changes, such as eye reshaping and weight changes. New guidelines allow ChatGpt to meet such requirements and change what is considered acceptable.
Openai is implementing these updates as part of its general mission of enhancing ChatGPT performance while also ensuring that more requests are processed without rejection. Openai maintains additional security measures for topics that we consider to be sensitive, including images of children.
Openai changed this policy during the expansion of public debate on AI content moderation to promote accountability for technology companies regarding moderation practices. Openai argues that policy revisions are irrelevant to political considerations, but these changes will spark new tests on ethical challenges regarding the creation of AI content and its implications for social debate.
In addition to relaxing content restrictions, Openai launched an upgraded image generator within ChatGPT, which attracted the attention of the internet. New features with the GPT-4O allow users to generate highly detailed images in a variety of artistic styles, including Studio Ghibli-inspired illustrations. The generator’s ability to create stunning pastel-colored artwork that reflects the aesthetics of a beloved Japanese animation studio has sparked excitement and excitement among users.
