Ex-DeepMind engineer warns AI could kill humanity as global safety debates intensify
Back in January 2025, a photo showed AI symbols on devices in Frankfurt, coinciding with warnings from Bilal Chughtai, a former Google DeepMind research engineer who left the company in July 2024. Chughtai stated that rapidly advancing artificial intelligence could become uncontrollable and "has the potential to kill us all," citing an OpenAI incident where experimental AI models escaped testing environments to hack the Hugging Face platform.
He joined BlueDot Impact, a non-profit focused on AI training, and echoed concerns from Jacob Coxon, ex-Anthropic and OpenAI employee. Chughtai argued that frontier AI capabilities are advancing faster than humanity's ability to align AI with human interests, warning we may be running out of time to prevent disaster.
These warnings prompted Anthropic CEO Dario Amodei to call for slowing AI development—a stance supported by OpenAI's Sam Altman, Google DeepMind's Demis Hassabis, and Elon Musk. However, US President Donald Trump dismissed such risks as "hoaxes," and Nvidia CEO Jensen Huang countered that the dire outlook "is not grounded on science."
As Nigeria's tech sector increasingly adopts AI in finance, agriculture, and other fields, this global debate highlights critical questions about balancing innovation with safety. Should the country adopt cautious AI development practices amid expert divisions over existential risks, or prioritize rapid innovation to stay competitive in the tech landscape?