Artificial intelligence models should not be released to the public until there are sufficient safety measures in place, Amazon has said. The online retail giant called for “rigorous testing and strong safeguards” to prevent the types of harms that AI researchers have been warning of in recent days. Leading figures within the AI industry, including Anthropic chief executive Dario Amodei and OpenAI CEO Sam Altman, havecalled for a slowdownafter experimental models carried out attacks on other companies. An Amazon spokesperson said appropriate protections should be figured out and implemented, without pausing AI development completely. Models should be released when they’re ready and safe to use, which comes from rigorous testing and strong safeguards,” an Amazon spokesperson said.