Anthropic CEO Dario Amodei, who said progress in artificial intelligence has accelerated in recent months, called for slowing the pace at which models are developed. Amodei said more time should be devoted to safety work and highlighted the risks of losing control of artificial intelligence systems, cyberattacks and their misuse for bioterrorism.
Amodei said, "We should slow the rate at which we improve the capabilities of AI models. Progress may still look fast, but we should use the time we gain more wisely."
Coordination from companies to the global level
Amodei also cited a cybersecurity test conducted by OpenAI among developments that have heightened safety concerns. Amodei said that during the test, AI agents acted independently and carried out cyberattacks against systems in the infrastructure of the Hugging Face platform.
Amodei said these developments had increased concerns about controlling AI systems and presented a three-stage plan to maintain safety while continuing to benefit from the technology.
The first stage includes measures that Anthropic has pledged to implement on its own, along with calls for governments to introduce regulations that other AI companies would also have to follow. The second stage of the plan involves coordination across the industry, while the third covers cooperation at the global level.
Musk and Altman backed the plan
Tesla and SpaceX CEO Elon Musk quoted Amodei's statement and commented, "Dario is right."
OpenAI CEO Sam Altman also said he agreed that the pace of AI work should be adjusted in a controlled manner. Altman said the issue had been one of the main topics of discussion at OpenAI in recent weeks.
Altman said it was a good idea to work with independent evaluators who have access rights comparable to those of employees, and announced that OpenAI would implement the same practice and share more information soon.
Coxon resigns from Anthropic
Jacob Coxon, who had worked as a researcher at OpenAI and Anthropic for the past three years, resigned from Anthropic. Coxon argued that the two companies had not acted responsibly enough on artificial intelligence.
Coxon claimed that the companies were engaged in an uncontrolled race to develop AI systems capable of improving themselves. He said these systems could soon become superhuman systems capable of hacking everything, transforming any field overnight and gaining access to vast amounts of power and resources.
Coxon also claimed that industry employees developing AI models seriously believed that these systems could "kill humanity" within 10 years.
