Archive · 14 Sep 2026
Elon Musk Renews Warning That AI Could Endanger Humanity
Elon Musk shared an old video about the dangers of artificial intelligence.
He said there is a small chance that AI could eventually harm or kill humanity.
Musk believes AI should be trained to tell the truth and remain curious.
He also said AI should be honest even when people dislike the answer.
His warning comes as some AI researchers worry that the technology is becoming very powerful very quickly.
Anthropic CEO Dario Amodei said fast-changing AI could be useful in healthcare but dangerous if not controlled.
OpenAI’s CEO agreed that AI development should move at a careful pace.
The discussion shows that people see both helpful possibilities and serious risks in advanced AI.
Elon Musk reshared a three-year-old video warning that AI might have a small chance of killing humanity.
He said AI should be trained to be truthful, curious and honest even when the truth is unpopular.
Musk’s post came amid renewed concern about artificial general intelligence and its potential existential risks.
Former AI safety researcher Jacob Coxon said AI developers believe it could kill people by the end of the decade.
Anthropic’s Dario Amodei and OpenAI’s CEO have called for slowing or carefully managing the pace of AI development.
- Who
- Elon Musk, Anthropic CEO Dario Amodei, OpenAI’s CEO and former AI safety researcher Jacob Coxon.
- What
- Musk renewed warnings that artificial intelligence could pose an existential risk and urged developers to prioritize truthful, curious and honest AI.
- When
- Musk shared the three-year-old video amid current concerns about rapidly advancing AI.
- Where
- The statements were shared online, including through posts on X.
- Why
- The warnings reflect concerns that advanced AI could become more capable than humans and potentially threaten humanity if its development is not carefully managed.
- Main warning
- Musk said AI has a small chance of killing all humans.
- AI qualities urged by Musk
- Truthfulness, maximum curiosity and honesty even when the truth is unpopular.
- Risk discussed
- Advanced artificial general intelligence could create existential risks for humanity.
- Researcher cited
- Jacob Coxon, who has worked in AI safety at OpenAI and Anthropic.
- Anthropic position
- Dario Amodei warned that rapidly adapting AI capabilities could become dangerous if not kept in check.
- OpenAI position
- OpenAI’s CEO agreed that the pace of AI development should be managed and supported independent evaluators with employee-like access.
- Potential benefit
- Amodei said rapidly advancing AI could help the healthcare sector.