Sam Altman Warns of AI Risks Amid Deepfake Video Surge

OpenAI CEO Sam Altman has sounded a serious warning about the rapid development of artificial intelligence, highlighting risks that could have far-reaching societal impacts. Speaking on the a16z podcast, Altman painted a picture of a future where AI could produce “really strange or scary moments,” if not carefully managed.

The caution comes as OpenAI’s new video application, Sora 2, gains attention online. Launched last month, the app allows users to create highly realistic video content, but early instances quickly surfaced of deepfake videos featuring famous figures, including a controversial portrayal of Dr. Martin Luther King Jr. This prompted OpenAI to issue a public apology and underline the challenges of managing AI content responsibly.

Altman emphasized the unprecedented speed at which AI video technology is reaching the mainstream. “Very soon the world is going to have to contend with incredible video models that can deepfake anyone or kind of show anything you want,” he said. The CEO believes society must “co-evolve” with AI, rather than releasing powerful tools without safeguards.

The discussion also touched on AI’s potential influence over human decisions. Altman warned of a future where many people might rely heavily on computers to make choices on their behalf, possibly triggering unexpected chain reactions at a societal scale. He illustrated this by imagining “billions of people talking to the same brain,” a scenario that could amplify risks in unpredictable ways.

Despite the warnings, Altman remains cautious about formal government regulation, suggesting that strict rules could create downsides. Instead, he advocates for “very careful safety testing” of AI systems, especially those that may become “extremely superhuman.” He concluded by stressing the importance of societal adaptation, believing that over time, communities will develop the necessary guardrails to navigate this rapidly evolving technology safely.

The episode underlines both the promise and the peril of AI. While innovations like Sora 2 demonstrate technological progress, they also highlight the urgent need for awareness, responsibility, and measured approaches to ensure AI’s benefits do not come at the cost of societal stability.