Cybersecurity professionals see risk everywhere, but partialities like salience bias, loss aversion, and target fixation can make us miss what matters, especially with AI. This talk will demonstrate how to spot these biases, set a realistic AI risk appetite, and shift from avoidance to optimisation.