Anthropic CEO urges slowing down AI development amid safety concerns

Dario Amodei, CEO of Anthropic, has called for a slowdown in the development of artificial intelligence models. He pointed out that AI has advanced drastically faster, including its ability to build the next generation of AI, raising concerns about safety and alignment.

In his essay titled We Must Pace the Frontier, Amodei emphasized the rapid progress in AI capabilities. He warned that the pace of development needs to be balanced to ensure safety while still achieving the benefits of AI technology. The proposal to slow down AI development has gained support from competitors who favor third-party monitors to evaluate the safety of models as they are developed.

Amodei suggested that if slowing down bought even an extra year or two before models reach critical levels of capability, and that time was used to advance alignment, the risk of something going seriously wrong could be greatly reduced. He stressed the importance of companies taking adequate time to align and safeguard their models, with third-party evaluators confirming this process.

The call for a measured approach does not imply halting model training or technical progress but aims to make AI safer by increasing transparency. Amodei also highlighted the need for voluntary cooperation among companies to set standards and called on governments to support these efforts.

In a related interview, Sam Altman expressed similar safety concerns, underscoring the importance of standards in AI development. The discussion reflects growing awareness of potential risks, with some experts suggesting there is a greater than 10% chance of serious issues arising if AI development continues unchecked.

The debate over pacing AI progress comes amid increasing competition in the field, with leaders advocating for a collective approach to ensure that advancements do not outpace safety measures.