Amazon enters AI safety fray, calls for "rigorous testing," safeguards

Amazon has entered the broader industry debate regarding artificial intelligence safety, advocating that AI models should undergo rigorous testing and maintain strong safeguards before public release. However, the tech giant stopped short of backing calls from rival AI laboratories for a deliberate slowdown in model development pace, known as pacing. The company stated that it does not view progress and safety as mutually exclusive, emphasizing that industry collaboration with government regulators is key to navigating potential risks. Amazon's comments come as it develops proprietary AI and artificial general intelligence models via Amazon Web Services, while also operating as a critical infrastructure and marketplace provider through its Bedrock platform. The statement highlights growing divergence among Big Tech firms regarding AI safety and regulation. While executives at Anthropic, OpenAI, DeepMind, and Microsoft recently backed a more cautious approach amid existential risk concerns, Amazon alongside Meta and Nvidia has favored private sector accountability over halting the pace of technological innovation.