Do No Harm: Unlocking Ethical Potential in AI Design - CyberEthicsLab