Research but don't deploy. There are enough unknowns to favor prudence; study risks, ethics, and governance before stepping into climate engineering.
