Extremely Important Cell Phone Feature Missing

it may require launching new hardware into orbit

I don't think so. It would require more hardware on the phones because they don't normally have the range. But the existing satellite networks definitely have the capability for this. Since this would be for something like expensive, rare use, the bandwith use would be minimal. One could use the Iridium satellite network easily.

However, I'm not at all sure that this is a serious existential risk threat except in so far as any general stability increase is likely to reduce existential risk.

Practically, this is unlikely to occur. People are unlikely to request this sort of thing (people don't think generally about low risk situations) although it might be possible to spin it in the context of the recent unrest in the Middle East. But most people aren't going to want a phone feature that only is going to be used if absolutely everything is going to shit. And without enough people interested in it, there's now way the cell phone manufacturers are going to bother.

I do not believe that making phones capable of contacting satellites, even just for texting, would be technically or economically simple; if it were, it would probably be at least an optional feature on many high-end phones. For developing countries in particular (to which most of your scenarios apply), the extra cost may very well be prohibitive.

First World users could probably afford a moderate increase in cellphone prices, but in the First World it's entirely possible to never leave regular coverage for extended time periods; and when you do, it's likely to either be in contexts wheres you would bring a separate GPS transmitter of some sort (extended camping / climbing trips, seafaring) or because you're somehow shielded from cell towers, in which case I don't know if satellite phones fare much better.

Many lives depend on this getting done.

As far as I know, a proposal that involves launching several new satellites will cost on the order of hundreds of millions of dollars (please correct me if the relevant hardware is cheaper in this case). As far as I know, hundreds of millions of dollars of medical spending even in some of the developed world can save thousands of lives.

So you really need to demonstrate the expected number of lives saved would be quite large. This seems to depend on some very strong claims about the availability of cell service. Do you have estimates of the number of people unserved by existing infrastructure, or who are unlikely to be served by infrastructure in the near future? Is there some reason that infrastructure will be unavailable in the case of disaster?

This could be the difference between a genocide going unreported, and being deterred.

This seems unlikely at face value. I have never heard of a situation in the modern world where there was complete information scarcity, such that isolated individual phone calls would significantly improve the situation. I have never seen any evidence that isolated reports of genocide have any positive effect as a deterrent. Could you describe in more detail the sort of situation you envision, or point to the evidence that led you to make this claim? (I am quite ignorant about these issues, so could be easily convinced by learning more.)

The genocide avoidance angle is novel and interesting, and it clearly seems like a good idea, but I don't get the existential risk angle.

I think the issue here is that connecting to satellite would require much greater signal strength, a different antenna design and a chipset for each satellite network you could contact (provided they each use a different coding and multiplexing scheme). You'd also probably need a handoff protocol of some kind, which could get really dicey. Basically, you'd be running into the same problems which prevent multi-network cellphones now, only far more aggressively plus have a range issue on top.

A simpler improvement would be to allow the cellphone to act as a powerful beacon. Basically, when the cellphone is put into emergency mode, it will transmit a strong signal every so often but will otherwise conserve power. This would solve the basic distress function very well.

An alternative: David Brin has proposed that cell phones should be capable of peer-to-peer text messaging when they can't connect with a tower.

(My father, a professor of electrical engineering, says that this proposal would be technically difficult to implement because cell phones transmit and receive on different frequencies.)

Satellite phones can currently be purchased by individuals or towns, and yet even the most affluent towns that fear being attacked don't appear to be purchasing them. A nonprofit could be set up to (in increasing order of importance)

1. Donate satellite phones to more impoverished towns.
2. Be a centralized location to which emergency messages can be sent.
3. Actually do something about genocide beyond issuing a press release.

My guess is that 3. is the sticking point - that a satellite capability on cell phones is actually only helpful to lost Westerners outside the range of a cellular tower, and might be best marketed to hikers.

An even lower cost universal solution might involve green lasers and Morse code, though various governments will object.

I think it is obvious that the desirability of this proposal is not obvious to an educated observer. If you believe the proposal important, the most important first step is not having connections (at least not to the political apparatus) but creating and presenting a compelling argument for its desirability.

For example: provide or reference more precise accounts of recent unrest, and present evidence that complete blackouts are a serious possibility. Provide evidence, or at least tell us why you believe, that receiving isolated communications could initiate a productive global response, or disincentivize tyranny. Provide evidence either that entire populations are unlikely to have access to traditional communication channels in the near future, or that a despot could cut off such channels without immediately causing much more alarm than any number of distress calls. Consider solutions that cost less than $100M, and conclude that they are not as cost-effective or not as easily implemented by government intervention. Etc.

In general, it seems to me that the world would be in a better state if we as a society had dismissed all proposals of the form "We should require providers of X to also provide Y" which didn't have very, very persuasive justification. So, taking the outside view, I am incredibly skeptical of any proposed regulation of this type.

This is pretty wrong. Cell phones that can talk to satellites do exist. They are expensive and people don't feel like buying them.

Mobile cell phone base stations can be moved around and put where needed. If you have an area with people in it that can't communicate, then already we have the abiltiy to send other people in with cell base stations and set them up. If people can't get to a population that has lost touch to see what's up, it means they can't or won't. "Falling off the grid" is enough of a cry for help already for a terrestrial population.

Cell is not the only way to communicate. Plenty of long distance radio communications exist. For a few hundred bucks you can buy a ham radio that will let you communicate 1/4 of the way around the world without breaking a sweat. If you think this is important, buy one and keep it in your safe room with your 3 years of canned food, bottled water, generators, rifles and dieself fuel. Whatever happens you will have a superb chance of reaching another ham operator out of your area and telling giving her the 411 on your 911.

The point of cell is CHEAP and UBIQUITOUS. Saying every cell should be able to reach a satellite in case of emergency is like saying every car should be able to do 200 mph and carry 5000 pounds of freight. It is not an idea that anybody who understands economics or engineering would resonate with.