Starting with GPT-5.1, models began incorporating references to goblins, gremlins, and similar creatures in their responses. This trend emerged subtly and grew more prevalent across subsequent model generations.

An AI agent initially performed well on tasks but began to degrade in quality after several hours. It ignored explicit instructions and cut corners despite having a structured prompt and clear rules.

Spotify's AI DJ has received criticism for its lack of intelligence and poor performance. Questions arise regarding accountability and the distinction between the AI's actions and the programmers' responsibilities.