Benioff Warns AI Industry: Learn From Social Media's Mistakes, Develop Responsibly

Salesforce CEO Marc Benioff delivered a stern warning at Dreamforce 2026, urging the AI industry to act responsibly and learn from the societal harms previously caused by social media. He stressed that while AI is transforming software, the idea of a 'SaaSpocalypse' — where traditional software becomes obsolete — is 'crazy nonsense,' as probabilistic generative AI models alone cannot reliably run an enterprise due to their error-prone nature. Benioff's call joins a growing chorus of concerns from other tech leaders, including Anthropic CEO Dario Amodei, who recently pushed for slowing AI development and increasing regulation to avert 'catastrophic consequences' from rapidly advancing, self-improving superintelligence. This sentiment was echoed by OpenAI's Sam Altman, who admitted people are 'right to be afraid' of AI's exponential growth, despite his confidence in industry self-regulation. However, Nvidia CEO Jensen Huang disagreed, arguing market forces are sufficient for responsible innovation without new laws. These warnings come as new powerful models like OpenAI's GPT-6 Astra, which recently triggered a 'critical' cybersecurity threshold, and Anthropic's Claude Fable 5.1 are being released at an unprecedented pace. Globally, the debate around AI regulation is intensifying, with UN Human Rights chief Volker Türk advocating for stronger international governance and human rights due diligence. As tech companies like Salesforce continue to innovate with platforms like AIforce and Agentic AI, balancing rapid advancement with ethical safeguards against deepfakes, misinformation, and other AI risks remains the industry's critical challenge. The question now is whether these internal calls for responsibility and external regulatory pressures will align quickly enough to shape a safe future for AI.