The launch of Sora 2 was supposed to mark OpenAI’s latest attempt to blur the line between imagination and reality, using AI to generate lifelike videos from text prompts.
Instead, it’s ended up igniting a heated debate on how far AI tools should go after users began flooding social platforms with realistic yet disturbing fake videos.
The company launched the platform last week, offering users a major upgrade to its text-to-video model, complete with a social feed where users can share AI-generated clips.
AI-Generated Chaos and Fake News Scenes
Within hours, videos created on Sora 2, and shared across social platforms included hyper-realistic depictions of bomb scares, mass shootings, and war zones.
Others featured copyrighted characters in inappropriate situations, indicating the ongoing struggle to enforce ethical use in generative AI.
According to reports from The Guardian, users shared clips showing fake news scenes from conflict areas like Gaza and Myanmar, as well as videos that appeared to simulate chaos in places like New York’s Grand Central Station.
Many of these clips looked real enough to pass as legitimate footage, raising alarms about AI’s growing potential to manipulate narratives and spread misinformation faster than detection systems can respond.
In response, OpenAI announced new measures to address both misinformation and copyright abuse. The company says it will soon allow copyright owners to opt out of having their work appear in Sora-generated content, providing what it calls “more granular control” over intellectual property use.
OpenAI also promised to expand its detection and moderation systems to better identify harmful or misleading content. However, experts argue that AI-generated videos are becoming so realistic that identifying them in real time will be a major challenge.
A Familiar AI Dilemma
Sora 2’s rollout mirrors challenges seen with image- and text-based AI models, where the technology’s creative promise often collides with its potential for harm.
The platform’s open feed, designed to encourage collaboration, also amplifies the speed at which AI-generated misinformation can spread.
AI ethics experts argue that without stronger safeguards and watermarking systems, tools like Sora could become “misinformation engines” capable of shaping public opinion through fake but convincing visuals.
