Suhas Subramanyam

Not analysed yet

A requested analysis is usually ready within half an hour; come back to this page or reload it.

Mr. Speaker, I sat down with some of the leaders in the AI industry. They told me that there were going to be dire consequences if we didn't put AI safeguards in place and that there would be massive implications for our country and for the world if we didn't act. That meeting was 10 years ago. I was in the White House as a technology policy adviser, and the AI leaders were saying then that we needed to act now. Here we are today. We have AI frontier models with the capability to wreak havoc. We are getting to a place where AI can self-improve and duplicate itself. The stuff that was theoretical 10 years ago is now happening. We have swarms of AI chatbots working together in a Machiavellian way to hack and undermine systems. We must act now. There are commonsense ways that security researchers, industry experts, and even Members of Congress here agree can make this safer, like auditing, information reporting, transparency, building safety throughout the AI development process, and real containment strategies, so that AI models don't escape their sandbox during development. This is common sense, and we need to act now. My colleagues and I have already put forward bipartisan legislation, like the FRONTIER Act, led by Representatives Obernolte and Trahan, and other bills. It is because voluntary standards are not a true safeguard, and sitting on the sidelines is not a safeguard. Even President Trump, despite what he just said online, is not a safeguard for the rapid development of AI and what it can cause. We need action now. We need to put legislation on the floor this week. We cannot wait until the next Congress.