You would need to consult a nation before enacting an idea, if this idea expressly involves them and your idea expects things of them. Prior to this consultation it is a poor idea as it bases on an unknown variable, which if accurately analysed doesn't yeild a particularly high probability of selection for acceptance by any right thinking nation.
You would also need to do a feasibility study on obtaining an island, setting up the prison infrastructure, the supporting infrastructure, and the applicable running costs for. Doing this on an island you also expect to be habitating 200 people on, you'd also need to add the costs of the feasibility study for that into account.
The joke is most definitely on you, buying an island is not even ½ the battle, one that doens't flood regularly with the tide and a large enough expanse to build for 200 heads, and associated infrastructure, isn't going to be cheap. Then it will cost possibly double that in getting the materials there, half of that again on materials. This doesn't buy architechts, civil engineers, project managers, builders, labourers etc. You exceeded the cost of my suggestion before obtaining the island. Vastly. And this mitigates the cost of building a prison in space, when you have a nigh-on infinite stack of resources of all types, anything you want to do is just how long it takes to do it. Then you don't have to try and turn it into a tool of generating revenue and it can stay true to it's purpose instead of corrupting itself and all that pass through by recieve more harm than good as a result. Things like maintainence costs are not really a thing, as generally automation suggests for the most part it should take care of itself. You don't spend money automating this, as similar systems from other facilities will be readily repurposed, as is the point of designing in a modular fashion. And another feather in the hat for open source development. You only waste effort doing something once, then snap together like lego. When it comes to automating a house, if you know what you're doing it's actually remarkably simple and can be done with COTS hardware reasonably cheaply. You can build it all from scratch for cheaper than you'd imagine too.
I don't listen to text, your way of thinking is far more dangerous than you seem capable of comprehending or your narrow scope of focus allows for. You may consider me "idealistic" - suggestion of this to most who actually know me should be interesting for you to receive their mockery, something else that should require youtubing for prosperity - I can assure you few have a pessimistic streak wider or deeper than mine. Where you see this discarded it is because all potential concerns have been ratified in this section. And if I cannot provide one, let alone multiple failure points - and I'm really good at breaking things, it's the only thing I do better than and what allows me to be so good at solving problems then it means there are viable solutions. Within reach. And a more than probable chance of succcess.
Your "way of thinking" does nothing to minimise loss of life, only increase dependance on revenue. My way of thinking most certainly is incredibly centric on prevention of loss of life, and has featured as core principle of several concepts. I'm quite up to date with methodologies, techniques, tools, and equipment currently in use associated with doing so and should be reasonably competent in at least outlining of sane mitigation strategies and techniques for almost any given situation - and ultimately this is likely to depend very little on my output in this regard. This concept of good design isn't an arrogant thumb in the nose of the icebergs of space as we destin ourselves to become the first space titanic, it's simple recognition that every "challenge" that exists can be met and overcome, as most already have been, with ease. With mitigation strategies in place to minimise loss of life in the unlikely event of failure, which are also consequence of good design.
When sat behind a five meter thick radiation shield, worrying about gamma rays that will only penetrate a few CM of a material as dense as NiFe, let alone mostly bounce off the 6 cm titaium skin on the outside, and never get near the one on the inside, it's highly irrational to anyone that is familiar with the likes of FLUKA and MCNPX. Or even someone that understands the basic working properties of the radiation they seem determined to be so scared of. There really isn't anything we can't do anything about, given sufficient preparatory time - and there's certainly no hope of building anything significant tomorrow. We have this time.
You've still yet to offer any valid evidence that an "unpaid workforce" would be possible to corrupt, or would be unwise for placement in any significant position. Surely there is some reason behind the almost religous finactical cling to such flawed concepts. You have certainly not offered any validation for why the act of attribution of monetary compensation for their priceless asset of time is any assurance of suitability for purpose in terms of competencies - existing systems commonly suggest otherwise - and why it magically provides prevention of them wishing to abuse their position to cause harm. Existing systems commonly prove otherwise there, too.
You don't smoke what I've got, and you can't buy it. It's all around you, you just have to open your eyes and look. Everything does work out as planned, or your plans were pretty unrealistic. Learn from this. I definitely don't seem to have problems with waking up, either, and you should find it disturbing this kid is vastly better educated.