AI Race May Endanger Humanity, Ex-Google Researcher Warns

A former Google DeepMind research engineer has raised fresh concerns over the potential dangers of advanced artificial intelligence (AI), warning that the technology could ultimately pose an existential threat to humanity if its development is not properly controlled.

Bilal Chughtai, who worked on artificial general intelligence (AGI) safety and alignment research at Google DeepMind before resigning in July 2026, said he believed AI had the potential to “kill us all” and warned that the world might be running out of time to prevent such an outcome.

Chughtai’s warning adds to a growing chorus of AI researchers and industry insiders expressing concern about the rapid development of increasingly capable systems and the possibility that safety measures may fail to keep pace with technological advancement.

Writing after he departed from Google DeepMind, Chughtai said AI could still be developed safely, but argued that this would require greater coordination among technology companies and an end to what he described as an intense race among AI developers.

“We need to pace AI development to a speed that society can handle, where emerging risks can be addressed before extreme harm is realised,” he said.

His comments came shortly after Anthropic researcher Jacob Coxon also resigned and warned that people developing advanced AI systems believe the technology could potentially kill humanity by the end of the decade.

Another Anthropic scientist, Evan Hubinger, subsequently said he agreed with the assessment and put the probability of AI killing all humans within the next decade at more than 10 per cent. The claims have intensified an already heated debate over whether the global AI race is moving faster than governments, researchers and safety institutions can adequately manage.

The concerns have also reached the highest levels of the AI industry. Anthropic chief executive officer Dario Amodei has called for a slowdown in developing the most advanced AI systems, arguing that the risks of increasingly powerful models require stronger safeguards.

The proposal has received support from OpenAI chief executive Sam Altman and SpaceX and xAI chief Elon Musk, while other technology leaders and governments remain divided over whether slowing AI development would create economic and strategic disadvantages.

Meanwhile, the United States is considering additional measures to compel major AI developers to demonstrate that their systems are adequately protected against catastrophic risks. US lawmakers are weighing proposals that could give government authorities greater powers to scrutinise AI safety measures and, in extreme circumstances, restrict the deployment of potentially dangerous models.

However, the push for tighter controls has faced opposition from some political leaders who argue that excessive regulation could weaken US competitiveness, particularly against China. US House Speaker Mike Johnson, for instance, has warned that a moratorium on AI development could hand China a strategic advantage.

Consequently, the debate has shifted beyond whether AI will transform economies and industries to a more fundamental question of whether humanity can maintain meaningful control over increasingly autonomous and powerful systems.

The latest warnings come as major AI companies continue to invest heavily in models capable of reasoning, operating autonomously and performing increasingly complex tasks. The growing concern among some researchers is that the systems may eventually develop capabilities that are difficult for their creators to predict, constrain or reverse.

However, Chughtai said the situation was not necessarily inevitable, maintaining that coordinated action and a more measured approach to AI development could help prevent catastrophic outcomes.

Hence, he underscores the widening divide between the drive to accelerate AI innovation and calls for stronger safeguards before the technology reaches more advanced stages.