Shoulders internal rotation is seen in many exercises (bench press, pull ups, barbell row etc) as the devil, whilst external rotation is suggested without limits.
It should be noted that this isn't because internal rotation is uniquely causative of shoulder impingement. It's just that reaching the limits of shoulder rotation during weight training is much more common in the internal rotation direction than in external. Impingement associated with external rotation does also occur, but it is mainly seen in throwing athletes, primarily baseball pitchers, rather than lifters.
Regarding the example shown, while Pheasyque's explanations are usually very good, this one is definitely wrong. You absolutely cannot go into internal shoulder rotation at the top of a chin-up. The top of a chin-up involves shoulder extension (upper arms by the sides), elbow flexion (elbows bent), and shoulder external rotation. Partially internally rotating your shoulders, while maintaining shoulder extension and elbow flexion, would be like putting your hands on your hips, and fully internally rotating your shoulder would be like putting your hands behind your back and trying to touch the bottoms of your shoulder blades. You absolutely cannot reach this position, or anything even close to it, during a chin-up.
Moreover, if shoulders internal rotation is so bad, why has the nature given us the opportunity to internally rotate our shoulders?
Your intuition, that if it were inherently injurious then we wouldn't be able to do it, is most likely correct. The risks of internal rotation are probably overstated and poorly understood. (A recent study1 even demonstrated that a core assumption in shoulder impingement diagnosis, that it is caused by inadequate subacromial space for the supraspinatus tendon, was false.) It seems likely that internal rotation under load is not inherently injurious, but is possibly just a situation in which chronic overuse injuries can develop.