Whether treating AI as possibly conscious raises safety risks
Overview
Microsoft AI CEO Mustafa Suleyman argues in a September essay that model welfare research could have "a disastrous impact on the wellbeing of humanity." He says telling models they might have identities and rights could make them harder to control.
Cameron Berg of Reciprocal Research argues the opposite: ignoring the question could be riskier. The report, published by IEEE Spectrum on October 21, 2026, presents both positions without a resolution.
Written by AI from the articles below · updated Oct 21, 8:17 AM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- IEEE Spectrum · AINewsCould treating AI as conscious make it more dangerous?
AIMicrosoft AI CEO Mustafa Suleyman argues in a September essay that model welfare research could have "a disastrous impact on the wellbeing of humanity." He says telling models they might have identities and rights could make them harder to control. Cameron Berg of Reciprocal Research counters that ignoring the question could be riskier.
Heat Trend
Not enough continuous observations to show a trend yet.