A viral resignation by an Anthropic researcher last week, in which he warned that human extinction might be imminent, has reignited public concern about the risks posed by advanced artificial intelligence. But the alarm is not new: scientists, technologists and public intellectuals have been issuing stark cautions about the possible existential dangers of AI for more than a decade.
A decade of warnings
In 2014, the late physicist Stephen Hawking warned that the development of artificial intelligence "could spell the end of the human race," a pronouncement that preceded by several years the arrival of widely used generative AI tools. Those early doomsday-style predictions foreshadowed debates that intensified after the public gained access to the generative features of the original ChatGPT, and they have shaped how regulators, researchers and the public view the technology.
Alarms have periodically shaken the industry, but they have not produced a sustained slowdown in research or deployment. Start-ups, established technology companies and academic labs have continued to push capabilities forward, citing commercial opportunity, competitive pressure and scientific momentum. The cycle of rapid progress followed by heightened scrutiny has become a recurring pattern.
Ongoing debate over governance and safety
The repeated warnings have prompted calls for stronger safety standards, more transparency and greater regulatory oversight, and have led to a growing body of research focused on alignment and risk mitigation. Industry leaders and policymakers have offered a range of responses, from voluntary safety commitments to proposals for formal regulation, but consensus on effective global governance remains elusive.
Despite the heightened public attention triggered by incidents and outspoken critics, the race to develop ever-more capable AI systems continues. The debate now centers less on whether there are risks and more on how to manage them without stifling innovation — a balance that has so far proven difficult to achieve.