OpenAI’s New ChatGPT Image Policy: Is AI Moderation Becoming Too Lax?

TL;DR – What You Need to Know in 30 Seconds

• ChatGPT can now generate images of public figures, previously disallowed.
• Requests related to physical and racial traits are now accepted.
• Controversial symbols are permitted in strictly educational contexts.
• OpenAI argues for nuanced moderation rather than blanket censorship.
• Move aligns with industry trends towards relaxed content moderation policies.

Is AI Moderation Becoming Too Lax?

ChatGPT just got a visual upgrade—generating whimsical Studio Ghibli-style images that quickly became an internet sensation. But look beyond these charming animations, and you’ll see something far more controversial: OpenAI has significantly eased its moderation policies, allowing users to generate images previously considered taboo. So, is this a timely move towards creative freedom or a risky step into a moderation minefield?

ChatGPT’s new visual prowess

OpenAI’s latest model, GPT-4o, introduces impressive image-generation capabilities directly inside ChatGPT. With advanced photo editing, sharper text rendering, and improved spatial representation, ChatGPT now rivals specialised image AI tools.

But the buzz isn’t just about cartoonish visuals; it’s about OpenAI’s major shift on sensitive content moderation.

Moving beyond blanket bans

Previously, if you asked ChatGPT to generate an image featuring public figures—say Donald Trump or Elon Musk—it would simply refuse. Similarly, requests for hateful symbols or modifications highlighting racial characteristics (like “make this person’s eyes look more Asian”) were strictly off-limits.

No longer. Joanne Jang, OpenAI’s model behaviour lead, explained the shift clearly:

“We’re shifting from blanket refusals in sensitive areas to a more precise approach focused on preventing real-world harm. The goal is to embrace humility—recognising how much we don’t know, and positioning ourselves to adapt as we learn.”

In short, fewer instant rejections, more nuanced responses.

Exactly what’s allowed now?

With this update, ChatGPT can now depict public figures upon request, moving away from selectively policing celebrity imagery. OpenAI will allow individuals to opt-out if they don’t want AI-generated images of themselves—shifting control back to users.

Controversially, ChatGPT also now accepts previously prohibited requests related to sensitive physical traits, like ethnicity or body shape adjustments, sparking fresh debate around ethical AI usage.

Handling the hottest topics

OpenAI is cautiously permitting requests involving controversial symbols—like swastikas—but only in neutral or educational contexts, never endorsing harmful ideologies. GPT-4o also continues to enforce stringent protections, especially around images involving children, setting even tighter standards than its predecessor, DALL-E 3.

Yet, loosening moderation around sensitive imagery has inevitably reignited fierce debates over censorship, freedom of speech, and AI’s ethical responsibilities.

A strategic shift or political move?

OpenAI maintains these changes are non-political, emphasising instead their longstanding commitment to user autonomy. But the timing is provocative, coinciding with increasing regulatory pressure and scrutiny from politicians like Republican Congressman Jim Jordan, who recently challenged tech companies about perceived biases in AI moderation.

This relaxation of restrictions echoes similar moves by other tech giants—Meta and X have also dialled back content moderation after facing similar criticisms. AI image moderation, however, poses unique risks due to its potential for widespread misinformation and cultural distortion, as Google’s recent controversy over historically inaccurate Gemini images has demonstrated.

What’s next for AI moderation?

ChatGPT’s new creative freedom has delighted users, but the wider implications remain uncertain. While memes featuring beloved animation styles flood social media, this same freedom could enable the rapid spread of less harmless imagery. OpenAI’s balancing act could quickly draw regulatory attention—particularly under the Trump administration’s more critical stance towards tech censorship.

The big question now: Where exactly do we draw the line between creative freedom and responsible moderation?

Let us know your thoughts in the comments below!

You may also like:

• China’s Bold Move: Shaping Global AI Regulation with Watermarks
• China’s Bold Move: Shaping Global AI Regulation with Watermarks
• Or try ChatGPT now by tapping here.