Artificial intelligence is developing at a rapid pace, raising both excitement about its potential and serious concerns about what increasingly powerful systems could mean for society. Recent warnings from AI researchers and industry insiders have intensified the debate over whether advanced AI could eventually become difficult for humans to control.
Some experts are particularly concerned about the possibility of highly capable AI systems becoming autonomous enough to make decisions, carry out complex tasks and potentially improve their own abilities with limited human involvement. If such systems were given access to critical digital infrastructure or other powerful tools, failures or misuse could have consequences far beyond today’s AI-related problems.
Over the next decade, AI is expected to become increasingly capable in areas such as software development, scientific research, cybersecurity, education, healthcare and business. At the same time, experts warn that the technology could increase risks including cyberattacks, misinformation, manipulation, job disruption and the misuse of AI for biological or other dangerous activities.
The most extreme concern is the possibility of humans losing control over a future form of artificial superintelligence. Some researchers believe this could theoretically create an existential threat to humanity, while other experts argue that extinction scenarios remain highly uncertain and should not be presented as an inevitable outcome. The International AI Safety Report 2026 stresses that AI capabilities are advancing quickly, while evidence about some future risks and effective safeguards remains limited.
The debate has also reached the political level. US Senator Bernie Sanders has called for stronger government action on AI, arguing that decisions about the future of the technology should not be left solely to a small number of powerful companies and executives. Meanwhile, AI industry leaders and researchers are increasingly calling for stronger safety testing, independent oversight and international cooperation.
For now, there is no reliable evidence that AI will inevitably kill humanity within the next decade. However, the rapid development of increasingly autonomous systems means the risks are being taken more seriously by scientists, technology companies and governments. The next ten years could therefore determine whether advanced AI becomes one of humanity’s most powerful tools or a technology that develops faster than society can safely control.
