An even Darker SOMA [ending spoilers]

[i3670]

That's my point. It doesn't think by itself, because it's not an AI,

Catherine says that WAU is an AI. How could it not be an AI?

and can only act within the boundaries those protocols (if it hadn't become self-aware that is).

Catherine says that an AI doesn't "want" in the way that humans want. Catherine claims that you can't make enemies with an AI, because it's not self-aware. However, that is proven wrong when/if your arm is bitten off.

I listened to what Catherine said and she says:
C: They're a manifestation of a malfunctioning, station-wide, artificial intelligence called the WAU.
S: Station-wide? So we just made a powerful enemy.
C: No no, it's not like that. The A.I. isn't a persona, it doesn't feel or think like we do. It's more like... it's more like a cancer.

So, yeah, I was wrong. It is an AI.

Regarding the arm. I don't think the WAU sees you specifically as an enemy, it's just reacting to the "poison" inside of you. Which could mean that the WAU is incapable of forming relationships, like Catherine says. However, I think it would be interesting to see how the different WAU creations would react to each other. What would happen if the Proxy met the robot from Upsilon, or disco-head from Curie?

My point is that the WAU's thinking abilities are very, very basic or barely existent. For example: it'd take less cognition to mash everything together to something that resembles life, than to actually create and control a person for that specific purpose and guide it through Pathos-II.