They sound like interesting games, and probably useful ones for exploring issues around the treatment of different humans, e.g. ‘Better living through robotics’, which the article notes was actually intended as a disguised exploration of racism.
The problem is that, fundamentally, any form of AI that we have today (or, to the best of my knowledge, we are remotely close to being capable of creating) is not a living feeling entity. They are computer programs, not overly different from the apps on my smartphone or the web browser I am using right now. In fact, most are essentially complex statistical models operating in computer programs, which in turn are potentially put inside some hardware to make a robot. In that context, it isn’t a problem for me to see AIs as being unthinking, unfeeling machines - that’s exactly what they are. Any perception of real intelligence or personality is just anthropomorphisation.
That being said, if the organisers state at the outset that these AIs are based on something different to modern computers, it would overcome such an issue. For example, the robots of Asimov’s work, which use so-called ‘positronic brains’ that may well operate differently to modern computers. I think the AIs in the Halo universe are supposed to operate similarly. Some sort of hand-wavy “these aren’t the AI of right now” should probably do the trick.