United States of America unveiled a broad executive order aimed at curbing the risks and “seizing the promise” of artificial intelligence commonly known as AI.
This was an effort by President Joe Biden’s aim to tackle complicated issues surrounding the fast-moving technology that US lawmakers have been struggling to meaningfully address.
President Joe Biden will sign an executive order which outlines “the most sweeping actions ever taken to protect Americans from the potential risks of AI” and will marshal government agencies and industry to address issues like privacy, trustworthiness and safety, the White House said in a statement.
Results of safety testing and other critical information will be required to be shared with the U.S. government under the Cold War era Defense Production Act.
This will be required from Companies developing powerful AI systems that could pose a threat to national security, the economy or public health as said by White House.
The National Institute of Standards and Technology will be directed to create a new set of “rigorous standards” to test AI systems before they are released and agencies that fund life science projects will use these standards as a condition for grants to “protect against the risks of using AI to engineer dangerous biological materials.”
Biden aims at boosting consumer protections and tackling areas where AI might fuel discrimination, including providing guidance to landlords and benefits programs on how to prevent algorithmic discrimination, developing guidelines agencies can use to assess privacy issues in AI and crafting best practices for how AI can be used within the justice system.
White House said it is also tackling deep fakes and other ways AI fuels disinformation.
The White House Administration said the Department of Commerce will develop guidance to clearly authenticate and watermark AI-generated content that will be used on official government communications “to make it easy for Americans to know” what they see from the government is real.
Biden is expected to sign the order at an event on Monday.
The White House said the order marks the strongest attempt for a government to tackle AI.
The executive order builds on voluntary commitments the White House secured from industry leaders like ChatGPT maker OpenAI, Meta, Amazon and Alphabet in July.
These firms pledged to abide by a set of safeguards to mitigate the risks of AI but there was no mechanism through which to enforce participation nor meaningful punishment should they deviate.
The risks of AI have been a topic of discussion among regulators, civic leaders, technologists and lawmakers for years but has acquired a new sense of urgency with the release of generative AI systems like ChatGPT, which has been rapidly taken up by the public and can generate content like text, music, art, data and video from scratch.
It raises a host of legal, social and ethical issues that cannot be easily resolved, such as who owns the products produced from a system trained on the work of many others and how to prevent AI systems from replicating and reinforcing the biases and discrimination from the data they are trained on.
As companies jealously guard their products and the data used to design them, many claims of transparency have been treated with skepticism.