Signal Boss: AI Threat Is Not Doomsday, But Negligence That Nearly Sparked a War

Meredith Whittaker, CEO of the Signal Foundation, dismisses the AI doomsday narrative that often dominates headlines. According to her, the fear that artificial intelligence will turn on humans has no strong scientific basis. She instead highlights a more down-to-earth danger: the careless use of technology that can trigger real disasters.

Whittaker points to an incident reported by CNN, in which a chatbot used by US intelligence produced false information about a Chinese ship carrying nuclear material. The erroneous information was said to have put the US on alert for a military response. "We almost went to war with China," she told BBC Global Women. For her, a system that produces flawed intelligence and nearly drags two major powers to the brink of conflict does not deserve to be called superintelligent.

Signal itself is not a giant in the messaging app market. With an estimated 40–70 million users, it is far below WhatsApp, which reaches billions. However, the app is known for its commitment to protecting privacy: Signal's servers do not store sender metadata, making surveillance by outsiders or hackers more difficult. This radical approach is what often puts Signal at odds with government policy.

Whittaker's background explains her uncompromising stance. A literature graduate who joined Google in 2006 as a temporary worker without a computer science degree, she rose to prominence in 2017 when she learned about Project Maven—Google's secret contract with the US Department of Defense to build AI for drone targeting. She realized the project aimed to "compress the kill chain": shortening the gap between a machine flagging a target and the decision to fire, reducing human deliberation. Along with thousands of colleagues, she refused. "It will launder bad intelligence and bad data," she recalled.

A year later, Whittaker helped organize a walkout joined by around 20,000 Google employees. They protested the company's handling of various issues, including what they saw as a rushed adoption of technology. Google then dropped the Maven contract, which was later taken over by Palantir. Whittaker draws a straight line from Maven to the US missile strike on a school in Iran this year, which she says relied on that system and outdated intelligence that identified the school as a military target. "This is what we warned about in 2017," she said.

Palantir denies this. The company says there is no evidence its software contributed to the tragedy and asserts it is not responsible for the underlying data or intelligence deficiencies. Palantir's UK and Europe boss, Louis Mosley, previously told the BBC that "there is always a human in the loop, so there is always a human making the final decision". US Defense Secretary Pete Hegseth also stressed the military would never target civilian sites. An official investigation into the Minab strike has not been published.

On the other side, governments—including the UK—are increasingly demanding access to conversations in messaging apps, citing the need to protect women and children. At London Tech Week 2026, then-Prime Minister Keir Starmer said technology companies must detect and block nude images on children's devices, or face legislation. Whittaker calls that approach an "unprecedented regime of communication and information control" and believes children will still find ways around it. The UK Department for Science, Innovation and Technology confirmed it is preparing legislation.

If a country forced Signal to weaken its encryption, would the company leave? "Yes! Good God, I have a spine," she said. "That doesn't change based on mood. It doesn't change because of consultants." She stressed that rather than expose users who depend on Signal—sometimes in life-and-death situations—to surveillance that could be misused, "we will leave the market".

"It has to work for everyone or it doesn't work for anyone. This is a principle of basic rights. Those rights apply even to jerks." — Meredith Whittaker

Criticism of Signal comes from child protection organizations. Rani Govender of the NSPCC said the group receives reports of children who are groomed and then moved to end-to-end encrypted platforms, where they are forced to create sexual abuse material. "Technology companies are responsible for the design choices they make, and they cannot ignore the risks that arise for children," she said. However, Signal also has support from women's groups, especially in the Global South. Hera Hussain, founder of Chayn, calls Signal a "survival tool amid state surveillance and cyberattacks", while acknowledging that any platform can be misused.

Whittaker also responded to a case in Canada, where several police officers shared racist and misogynistic messages in a Signal group that leaked after one member forwarded them to the media. She argues that if those conversations had taken place in a kitchen and someone leaked them, would we question the existence of kitchens? For her, communication privacy protects many people from various dangers: abusive partners, misogynistic law enforcement officers with database access, or authoritarian governments that do not want educated women.

For women who want to enter the technology industry but feel intimidated, Whittaker advises them not to be cowards in the face of bullies. "Don't laugh at jokes that aren't funny," she said. She stressed the importance of courage in facing an environment that is not always friendly. Meanwhile, in Indonesia, a similar discourse is emerging: the government is drafting regulations on personal data protection and cybersecurity that often intersect with encryption. Will Indonesia follow the UK's lead in demanding access, or instead become an example of a jurisdiction that respects end-to-end encryption? In the coming years, this policy choice will determine whether its citizens can communicate without the shadow of surveillance.