While a global AI watchdog could help mitigate risks, it is unclear how effective it would be in practice. The challenge lies in getting buy-in from other nations and industries, as well as ensuring that the watchdog remains free from political influence.