Illustration by TIME; reference image courtesy of Nina Jankowicz, photograph by Kate Andreya

Nina Jankowicz had been advising governments, organizations, and tech companies around the world about the effects of disinformation and fake news on democracy and free speech for years before the Biden Administration tapped her to lead its Disinformation Governance Board, created in April 2022 under the Department of Homeland Security to counter misinformation. But hours into her appointment, the then 33-year-old became the target of a sustained disinformation campaign herself. Right-wing trolls on the internet waged continuous attacks that included allegations that she was transgender and infertile. Prominent Republicans and Fox News, mining conservative perceptions of leftist bias in tech, derisively dubbed the board a “ministry of truth.” Jankowicz eventually resigned, and the U.S. government disbanded the board just three weeks after it was first announced.

More from TIME

Yet the attacks continued. In June, she discovered three deepfake porn videos featuring her official U.S. portrait; they had been uploaded online nearly a year prior. The experience was “a little bit surreal,” Jankowicz says, not least because she was six months pregnant at the time. “It was also clear they were being used to train an AI model.”

That model targets women. A 2019 study found that 96% of deepfakes found online are pornographic, and 100% featured women without their consent. But as Jankowicz sifted through deepfake websites and forums, she noticed other alarming trends. The videos primarily featured women who had “drawn the public’s ire” simply by working in fields from politics to the performing arts. One forum even had a running list of celebrities and influencers who were on the cusp of turning 18. “That was the most chilling part for me,” she says. “It was just a really strategic attack to denigrate, undermine, and embarrass women online.”

To her, the harrowing experience has made one thing clear: “AI’s relationship with gender and women is adversarial,” Jankowicz says. The author of How to Be a Woman Online and How to Lose the Information War now spends much of her time urging governments to tackle the spread of disinformation, online abuse, and deepfake pornography urgently, amid broader discussions about regulating AI. Last September, she announced the launch of the Hypatia Project at the Centre for Information Resilience, which seeks to document and combat these online harms. Without action, she warns, AI will continue to be trained on content with “an overwhelmingly male understanding of the world” and, in turn, perpetrate “significant harms” to women.

More Must-Reads from TIME

Write to Astha Rajvanshi at

Sougwen Chung
Yann LeCun
Abeba Birhane
Audrey Tang