Signal's Chief: Real AI Risk Isn't Sci-Fi, It's Careless Use

When a chatbot nearly sparked a war with China by inventing information about a nuclear vessel, one tech leader saw it as proof that AI's biggest threat isn't what dominates headlines.

Meredith Whittaker, who runs the Signal Foundation behind the encrypted messaging app, has little patience for stories about AI turning against humanity. She calls these warnings "shiny ghost stories" not backed by science.

The real danger, she argues, is far more immediate. That alleged incident involving US intelligence shows how systems deployed without proper safeguards can create genuine crises. "A system that produces faulty intelligence and almost gets us into a war with China should not be considered super-intelligent by any standards," she told BBC Global Women.

Whittaker's journey to this position started unconventionally. She studied literature, joined Google as an entry-level temp in 2006 without a computer science degree, and rose to become one of Time magazine's most powerful people in AI.

Her defining moment came in 2017 when she learned about Project Maven, a secret Google contract to build AI for drone targeting. Understanding it would "compress the kill chain" between identifying and firing at targets, she pushed back publicly alongside thousands of colleagues. Google eventually dropped the contract.

Why This Inspires

Whittaker's story shows that protecting people from technology's risks doesn't require a computer science degree. It requires backbone and willingness to speak up when others won't.

She draws connections between that early warning and this year's US missile attack on a school in Iran, which she says relied on the system and outdated intelligence. Her point: the consequences she predicted in 2017 are materializing now.

Signal itself operates on principles of radical privacy protection. With an estimated 40 to 70 million users, it's tiny compared to WhatsApp's billions. But it hides who sends messages even from its own servers, making surveillance nearly impossible.

When asked if Signal would leave countries like the UK that demand weakened encryption to protect children, Whittaker's answer was immediate. "Yes! My God, I have a backbone," she said. The app is already blocked in Russia, China, and Iran.

Her stance draws criticism from child safety advocates who point to real grooming cases on encrypted platforms. But women's groups in the Global South, who rely on Signal against state surveillance, support her position.

Whittaker's message is simple: the AI risks worth worrying about aren't in science fiction, but in rushed deployments and careless implementation happening right now.