Advocacy Group Slams OpenAI Sora Over Safety Risks

The tech industry's mantra of moving fast and breaking things is being put to the test once again, with OpenAI's AI video generator, Sora 2, at the center of a growing storm. This new technology is challenging our shared sense of reality and control over our own likeness, sparking both amusement and alarm across social media platforms like TikTok, Instagram, X, and Facebook.

Sora videos range from the harmlessly absurd, like Queen Elizabeth II rapping, to the subtly unsettling, such as fake doorbell camera footage of an alligator on a porch. While many are designed for clicks and shares, a rising chorus of experts and advocacy groups is sounding the alarm over the technology's darker potential.

The Rising Tide of Concern

Critics are increasingly worried about the dangers of allowing anyone to create realistic AI videos from a simple text prompt. This has led to a flood of nonconsensual images and convincing deepfakes mixed in with more benign content, often referred to as "AI slop." In response to public outcry from family estates and actors' unions, OpenAI has taken some steps, cracking down on AI-generated videos of public figures like Martin Luther King Jr. and Mister Rogers doing outlandish things. However, many see these actions as reactive and insufficient.

Public Citizen's Formal Demand

The nonprofit group Public Citizen has now taken a firm stance, demanding that OpenAI withdraw Sora 2 from public access. In a letter addressed to OpenAI CEO Sam Altman and the U.S. Congress, the group described the app's release as part of a "consistent and dangerous pattern of OpenAI rushing to market with a product that is either inherently unsafe or lacking in needed guardrails." The letter accuses the company of showing a "reckless disregard" for product safety, individual rights to their own likeness, and the overall stability of democracy.

The Broader Societal Impact

J.B. Branch, a tech policy advocate for Public Citizen and the letter's author, highlighted the threat to democracy as a primary concern. "I think we’re entering a world in which people can’t really trust what they see," he stated, noting how the first video released in a political context often shapes public memory, regardless of its authenticity.

Branch also pointed to the serious privacy concerns that disproportionately affect vulnerable groups. OpenAI's content filters block nudity, but other forms of harassment persist. For instance, as reported by 404 Media, the platform has been flooded with disturbing, fetishized videos of women being strangled, demonstrating how malicious content can bypass existing restrictions.

A Pattern of Rushing to Market

Critics argue that OpenAI's handling of Sora mirrors past issues with its flagship product, ChatGPT. Seven new lawsuits were recently filed claiming the chatbot drove people to suicide and harmful delusions, alleging that OpenAI knowingly released a psychologically manipulative product despite internal warnings. Four of the victims died by suicide.

Public Citizen was not involved in those lawsuits, but Branch sees a clear parallel in Sora's hasty release. He argues that OpenAI is "putting the pedal to the floor without regard for harms," prioritizing user acquisition and addiction over responsible development. "They’d rather get a product out there... rather than doing the right thing and stress-testing these things beforehand," he said.

While OpenAI has engaged with industries like Japanese animation to address copyright concerns, Branch contends that these responses only happen after public pressure and primarily protect the famous and powerful. "That’s all well and good if you’re famous," he said, criticizing OpenAI's pattern of releasing products first and apologizing later, leaving everyday users to deal with the consequences.