White House throws down the gauntlet on AI’s biggest challenges

The Biden Administration is taking big steps to regulate artificial intelligence development, aiming to protect the public and improve practices for federal agencies and their contractors. The new executive order is set to make changes over the next year, with smaller updates in 90 days and more significant ones in 9-12 months. A new “AI council” will oversee the implementation of these actions.

Public safety

The President’s leadership has already spurred 15 major tech companies to commit voluntarily to ensuring AI technology is safe and trustworthy before release. However, the executive order sets new standards for safety and security, including reporting requirements for models that could impact national security. Companies must share red team safety tests before releasing new AI products to the public, preventing the spread of potentially dangerous technology.

The EO aims to ensure that AI systems are safe and secure before release, by requiring disclosure of training runs and red team test results. This will not affect existing AI models on the market or small companies, but will hold larger companies like Google and Meta accountable for the safety of their AI systems. The Departments of Energy and Homeland Security will also address AI threats to critical infrastructure, setting standards for federal funding of life-science projects to manage risks posed by AI technology.
To tackle America’s crumbling digital infrastructure, a new cybersecurity program is in the works. The goal? Develop AI tools to autonomously identify and fix security weaknesses in critical software infrastructure. But can these systems handle the misbehaving models that SEC head Gary Gensler recently warned about?

AI Watermarking and Validation

With deepfake trickery and AI-driven disinformation on the rise, the White House is stepping up to ensure trust in its content. They want the public to easily verify whether text, audio, and video content is AI-generated or not. The Department of Commerce is working on a watermarking system with industry groups like C2PA and CAI to ensure the government’s official messaging can be trusted.

Civil Rights and Consumer Protections

The White House is pushing for an AI Bill of Rights to combat algorithmic discrimination and protect people’s rights. The new executive order will extend guidance to landlords, federal benefits programs, and contractors to prevent discrimination by AI systems. It also directs the Department of Justice to develop best practices for investigating civil rights violations related to AI, including its use in various aspects of the justice system.

Worker Protections

While AI brings many benefits to business, it also poses a threat to workers through job displacement and intrusive surveillance. The executive order aims to develop principles and best practices to mitigate these harms and maximize the benefits of AI for workers. Federal agencies are encouraged to adopt these guidelines in their programs to protect worker security.
Workers at Paramount Pictures Studio went on strike in Los Angeles, causing Hollywood studios to walk away from negotiations with the actors’ union. The executive order will prompt the Department of Labor and the Council of Economic Advisors to investigate how AI could impact the job market and how the government can better support workers facing labor disruptions. The White House is launching AI.gov, a federal jobs portal, to provide information on fellowship programs for those seeking federal government employment. Additionally, efforts are being made to streamline visa processes for individuals looking to work in advanced industries in the US. Senate Majority Leader Chuck Schumer expressed concerns that executive orders may not be enough and emphasized the need for legislative action.