As the race to develop advanced artificial intelligence (AI) models continues, three of the world’s biggest AI platforms have called for a slowdown in model development due to escalating risks associated with the technology. Anthropic PBC CEO Dario Amodei has implemented new safety steps, including third-party evaluators, and has urged the industry to support a broader downshift. OpenAI chief Sam Altman has pledged to adopt Amodei’s suggestion of “independent evaluators with employee-like access,” while Elon Musk, who runs xAI Corp., has expressed his support for the move.

While the leaders have long warned about the risks of AI, this is the first time they have coordinated to slow down the development pace. The move comes as anxiety about severe AI risks enters the mainstream, driven in part by recent high-profile resignations within the industry and dire warnings from experts. The Chinese government has rejected these warnings, calling them “fearmongering” and accusing the US of engaging in “confrontation and vicious competition.”

The push for a slowdown in AI development comes as the technology continues to evolve at an unprecedented pace. As companies race to develop more advanced models, there is growing concern about the potential risks associated with these technologies. These include the possibility of AI systems causing harm if they are not designed with safety in mind, as well as the potential for AI to be used for malicious purposes.

The move by leading AI executives to slow down model development is a significant shift in an industry that has historically prioritized speed and innovation over safety and responsible development. It remains to be seen how far these companies will go in imposing new limits or safety checks, but the fact that they are taking steps to address these concerns is a positive sign for those who are concerned about the potential risks of AI. As the technology continues to evolve, it is crucial that developers prioritize safety and responsibility in their pursuit of innovation.