In today’s world, Artificial Intelligence (AI) systems pose numerous challenges, such as ineffective or unsafe patient care, biases in hiring procedures, undermining people’s privacy, or tracking their activities without their consent. Therefore, the White House Office of Science and Technology Policy has recently unveiled a legally non-binding document to better protect the American public in the AI age. The paper offers a set of guidelines that companies can follow to protect users better. It provides suggestions on how to incorporate AI rights in technical design.
The document outlines five protections with the use of AI
Safe and Effective Systems
Users should be protected from unsafe or ineffective systems and harm from unintended uses or impacts of automated systems. Different communities, stakeholders, and domain experts should be consulted before developing any automated system to identify its risks and potential effects. Evaluation and reporting should be done to confirm that the system is safe and effective, and the results should be made public.
Algorithmic Discrimination Protections
There should be safeguards against the differential treatment of people by algorithms based on their race, color, ethnicity, sex, religion, national origin, disability, or any other classification protected by law. Such discrimination includes lower credit scores, expensive mortgages, students getting screened out of better schools, etc.
Users should be protected from abusive data practices and violations of privacy. Users should have agency over the use of their data. The user’s permission regarding their data use, access, transfer, and deletion should be sought, and their decisions regarding the same should be respected. Any consent request should be brief and understandable in plain language.
Notice and Explanation
Users should be notified when an automated system is in use, and they should understand how they are affected by the decisions made by the AI system. The designers, developers, and deployers of such systems should provide easy-to-understand documents having clear descriptions of the overall system and the role which automation plays. Users should know how and why an automated system determines an outcome impacting them.
Human Alternatives, Consideration, and Fallback
Users should be able to opt-out of an automated system whenever they wish to. Moreover, they should also have access to someone who can resolve their issues quickly and an escalation process in case of failure of an AI system.
” The AI Bill of Rights could have a monumental impact on fundamental civil liberties for Black and Latino people across the nation.”
~ Willmary Escoto (US Policy Analyst for Access Now)
The AI Bill of Rights forms a set of protections against potential harm. It marks an essential step in recognizing how algorithmic systems can deepen inequality. It can impact the American public’s rights, opportunities, or access to critical resources or services. It empowers the American people to expect and demand better from their technologies, giving more control over their data and ensuring that automated systems are used safely and transparently.
The Bill, however, does not discuss the controversial forms of AI, such as mass surveillance using facial recognition or autonomous lethal weapons. Moreover, the principles stated in the Bill are not intended to and do not prohibit or limit any lawful activity of a government agency, including law enforcement, national security, or intelligence activities.
Over the past few years, many tech giants, like Facebook, Google, Alibaba, etc., have been under the scrutiny of their respective governments. The AI Bill of Rights is considered America’s latest effort to scrutinize the tech giants and hold them accountable. Different tech companies, however, like Palantir and Microsoft, were consulted, and their feedback was incorporated into the Bill. The technology sector has also welcomed the White House’s acknowledgment that Artificial Intelligence can also be used for good.
Non-enforceability is said to be the biggest issue with this document. It only acts as a guide for businesses developing and deploying AI systems. Companies can ignore it, which could blunt its impact. However, the administration has announced that various federal agencies would take steps to put the principles into practice.
The document is just a starting point to move the US to a place where it can carry forward on the commitment to implement human-centric and trustworthy AI. It forms an overlapping set of safeguards, and hopefully, it will help to protect the public from the dangers of AI systems.
Please Don't Forget To Join Our ML Subreddit
I am a Civil Engineering Graduate (2022) from Jamia Millia Islamia, New Delhi, and I have a keen interest in Data Science, especially Neural Networks and their application in various areas.