Dario Amodei, CEO of Anthropic and former OpenAI employee, has been scheduled to testify at a US Senate hearing on AI. The 'Oversight of AI: Principles for Regulation' hearing aims to explore potential regulations for the rapidly growing technology. In addition to Amodei, the hearing will also feature computer science professors Yoshua Bengio and Stuart Russell as witnesses. The Senate Judiciary Subcommittee on Privacy, Technology, and the Law will host the hearing on 25 July.

'It's our obligation to address AI's potential threats and risks before they become real,' stated Democratic Senator Richard Blumenthal, who chairs the subcommittee. Blumenthal underscored that there are significant consequences for workers, consumer privacy, and society as a whole in this upcoming era. Senator Josh Hawley, the leading Republican on the panel, shared similar concerns about the impact of AI on our country. He stressed the importance of safeguarding consumers, workers, and families from any potential negative effects caused by AI.

At the recent Fortune Brainstorm Tech conference, Dario Amodei shared his insights on AI, shedding light on several concerns that he categorised into three tiers. In the short term, issues like bias and misinformation were identified. Medium-term risks were related to the advancement of AI models in fields like science and engineering, which could potentially lead to harmful applications. In the long term, the emergence of models with agency and autonomy posed significant challenges in controlling their actions, potentially leading to existential threats. Amodei stressed the utmost importance of designing AI systems with safety as a primary consideration to mitigate any negative consequences.