Yes it did, which is the main reason why I prefer Civ 4. It was probably dumber in C4, but had better judgement.
The whole point of Civ5's AI is that it wants to win, it picks a victory condition early in the game and then pursues it relentlessly. In fact, that's why so many people hated the Civ5 AI, because if another civ got a hint that you were going for the same victory as it then they would hate you for it (something that was removed from G&K). Civ4's AI was more roleplay-y, it attempts to achieve victory but it's not as consumed with it as Civ5's. I just got beat to a spaceship victory in my most recent game and, let me tell you, the AI went for that victory with gusto. They completed the Apollo Program about 60 turns after me (marathon game), and then built all the components of the ship in the next 30 turns, while I was still at least 50 turns away from getting the last component. A few months ago I played a game with a real nail-biter of an ending, with four different civs vying for 4 different victories, which I only just barely won.
That's why I'm so confused by how Arabia flubbed their lead, in dozens of of Civ5 games I've never seen an AI screw up such a clear lead. What may have happened was that they were attempting a domination victory and had targeted me as the next victim as I was the weakest of the three remaining civs, but I had a defensive pact with Mongolia and Arabia was waiting for that to run its course before attacking me. But because diplomacy was reworked in G&K, things like defensive pacts can be renegotiated between turns so they never run out, and Arabia never never got an opportunity to take a shot at me without taking on Mongolia as well, which they weren't prepared to do. If that's the case then they should have gone for a plan B or C and aimed for a diplomatic or scientific victory, and that would indicate there's a problem with the way the AI pursues victory.