Maybe if several low power lasers were fired together, to converge at the predicted position, it would moderate the risk of collateral damage. Even a few millimeters away from the target position, the radiative flux (W/m²) would be much lower, likely at safe levels.

What if all of those lasers converge to hit someone's eye?

I don't see how a laser can have enough power density to kill a mosquito but still be eye safe.

Is there a max angle at which two lasers could pass through the pupil and strike the retina? That's assuming that strikes elsewhere don't pose a blinding risk for the energy level required to kill a mosquito, which I have absolutely no support for. But if that is the case, you could place your lasers and define a kill zone where no two can converge with less than a safe angle.

If the assumption isn't correct, you could possibly still place at least two lasers and define a narrow zone between them where the angle would be about 180 degrees. Would be damned tough to hit an eye from two sides at once. But that limits you to half power per laser, which probably isn't enough margin.

We are talking about two different safety issues:

(1) Targeting the wrong thing.

(2) Even if the targeting is correct, missing by a small amount, or running the laser for slightly too long, and hitting something else by mistake.

If the lasers converge for full flux, the risks associated with (2) are better managed, but that doesn't speak to (1).

IIRC (can't find a reference), the idea was noty to be able to kill teh mosquitoes outright, but to damage their wings so they couldn't fly. _Much_ less power needed.

Still an issue, though, but not so bad

Like my retinas.

Perhaps if the laser light were put through adjustable optics to set the focal point at the target. That way, anything sufficiently closer or further than the target along the laser's path will see non-dangerous intensity. Still don't want the thing accidentally targeting eye balls though.