OpenAI says its new model GPT-2 is too dangerous to release (2019)

OpenAI announced GPT-2, a sophisticated text-generation model trained on 8 million webpages, but withheld the full algorithm and training data citing safety concerns about potential misuse for generating disinformation, impersonation, and spam. The decision sparked debate within the AI research community about whether the risks were overstated versus OpenAI's role in raising ethical awareness around powerful AI systems.