Are the AI civs strength/power randomly generated?

Sometimes i noticed that an AI civ, Civ X for example, would perform extremely well in a game, but perform noticably weaker in another similar game.

For example, Egypt in one game had became the world power but in another, it was a third world. In both games Egypt had similar land size, neighbours, location and luxuries.
Also, for example, Brazil in one game had the biggest cultural output and the most wonders. In a similar game in a similar map, with similar(near similar) civs, it fails.
Also, sometimes why does a much smaller, little luxury civ perform so much better compared to a larger one?

Example, India in one game had very few luxuries and had one of the smallest land areas, but its tech was no. 1, reached modern era while others just started industrial, had the most GNP, Gold per turn, etc etc. It seems as though India was cheating because its size cannot possibly enable it to be so overpowered.

AI Civs performances are determined by several factors:

1) The layout of the land
2) Who their neighbors are
3) What their randomly determined personality tells them to do this game. (More on this in a bit)
4) RNG with ruins & barbarians

With regard to their personalities, all AI civs have base values (on a scale of 1 to 10) for various attributes, (e.g. aggression, tendency to sign DoF, etc.) but their specific values for any given game can vary by +-2.

In addition to all the above variables, they also seem to semi-randomly choose their social policy trees. If they end up choosing Piety (or to a lesser extent Honor) then there's a good chance they're not going to do well in that game.

It can really depend on difficulty level too.

India was rarely a problem for me when I played King, but on Emperor, they tend to be a problem next to a warmonger. Gandhi ends up beating the warmonger and taking him out of the game, being hated by everyone, and one by one the AI DoW's him and gets eaten by his expanding, well developed empire.

If he doesn't get DoW'ed early, he's a few free cities the rest of the game.

Siam can get out of control the other way, being near peaceful civs. He'll play nice for a while then DoW over anything later. At lower difficulties, he can't manage to take a city most of the time. At a higher difficulty, he'll have the units and production to start the steamroller.

I have always found Attila to do better on slightly lower difficulty. When I played Prince, he often had the ability to grow though war early more effectively now that I play on Emperor. The other AI's are more likely to keep a decent army and be able to stop his rams at higher difficulties, I guess, leading him to be a bit less threatening if left alone.