The White House just announced a thunderous executive order on artificial intelligence regulation. According to White House Deputy Chief of Staff Bruce Reed, these directives represent “the strongest set of actions ever taken by any government in the world” to protect the impact of artificial intelligence on American citizens.
The Biden administration is working on plans to regulate the independent artificial intelligence industry. The order builds on the Biden-Harris plan for AI Bill of Rights but also voluntary commitments of 15 leading technology companies to work with the government to develop artificial intelligence safely and responsibly.
Instead of waiting for Congress to pass its own legislation, the White House is storming through with an executive order aimed at mitigating the risks of artificial intelligence while realizing its potential. With the widespread operate of generative AI such as ChatGPT, the urgent need for AI is real.
White House AI Executive Order: 10 Key Provisions You Need to Know
What does an executive order look like? And how will this affect AI companies? Here’s what you need to know.
1. Developers of powerful AI systems (e.g. OpenAI, Google and Microsoft) must share the results of their security tests with the federal government
In other words, while a leading AI company is training its model, it is required to release the results of the red team’s security tests before are made public. (The red team is a group of people who test the security of a digital entity by posing as malicious actors.)
According to a senior administration official, the order focuses on future generation of AI models, rather than current consumer-facing tools like ChatGPT. Moreover, the companies that would be required to share security results are those that meet the highest computational performance threshold. “[The threshold] it will not cover artificial intelligence systems trained by students or even professors. “This really is about the most powerful systems in the world,” the official said.
2. Red team testing will be conducted to the high standards set by the National Institute of Standards and Technology
The Departments of Homeland Security and the Departments of Energy will also work together to determine whether artificial intelligence systems pose certain risks to cybersecurity as well as our chemical, biological, radiological and nuclear infrastructure.
3. Address AI security with models in science and biology projects
Work is underway on novel “biosynthetic screening” standards to protect against “hazardous biological materials” developed by artificial intelligence.
4. Content generated by artificial intelligence must be watermarked
The Department of Commerce will develop guidelines to ensure that all AI-generated content – audio, images, video and text – is labeled as such. This will enable Americans to determine which content is created by a non-human entity, making it easier to identify dishonest deepfakes.
5. Continue to operate the “AI Cyber Challenge”
For the uninitiated, the AI Cyber Challenge is a Biden administration initiative that aims to establish a high-level cybersecurity program that strengthens the security of artificial intelligence tools, ensuring security vulnerabilities are addressed.
6. Rely on Congress to Pass “Bipartisan Data Privacy Legislation”
The executive order is a message to Congress to speed up the work. Biden is calling on lawmakers to ensure Americans’ privacy is protected when leading players in artificial intelligence train their models. Children’s privacy will be a major topic.
7. Review company data policies.
The White House says it will assess how agencies and third-party data brokers collect and operate “commercially available” information, or public datasets. Some “personally identifiable” information is publicly available, but that doesn’t mean AI players can freely operate that information.
8. Fight discrimination exacerbated by AI
The guidance will be shared with property owners, federal contractors and others to reduce the potential for bias. In addition, the government will introduce best practices to counteract discrimination in artificial intelligence algorithms. Additionally, the Biden administration will address the operate of artificial intelligence in sentencing within the criminal justice system.
9. Attract the world’s best talent
From today ai.gov This website provides a portal for candidates seeking AI-related scholarships and employment opportunities with the US government. The order also aims to update visa criteria for immigrants with expertise in artificial intelligence.
10. Support employees susceptible to the development of artificial intelligence
The Biden administration will support the impact of worker collective bargaining by developing policies and best practices to protect workers from potential harms such as surveillance, job displacement and discrimination. The order also announced plans to produce a report on the potential of artificial intelligence to disrupt labor markets.
Mashable will be in Washington for more information on how the novel AI executive order will impact major players like Open AI, Google and Microsoft, as well as the average American citizen. Stay up to date with our coverage of this topic.
