Self Harm Prevention

OpenAI introduces new ‘Trusted Contact’ safeguard for cases of possible self-harm