MEMORANDUM
To: Interested Parties
From: Sacha Haworth, Executive Director, Tech Oversight Project
Re: Polling on Regulating Artificial Intelligence (AI) Technology
Date: November 29, 2023
PURPOSE: Congress needs to consider the full spectrum of AI harms, heed public polling, and pass urgently-needed AI protections that will protect competition, mitigate consumer harms, and support our democracy and national security right now.
As President Joe Biden laid out in his Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence, the risks of unregulated Artificial Intelligence are not solely hypothetical and in the future – AI is creating harm right now.
As Bloomberg reports, “a new era of disinformation is dawning. While experts have long warned about the use of deepfakes to sway voters, AI is now cheap and accessible enough for anyone to try their hands at it.”
- AI tech has facilitated discrimination in housing, criminal justice, and health care.
- AI is being used to scam elderly people by cloning voices of their loved ones and asking the elderly to send money. These are real threats, and immediate action must be taken.
- In recent elections in the EU, disinformation from AI deepfake audio and videos were used to sway the electorate.
President Biden’s Executive Order aims to establish “new standards for AI safety and security, protects Americans’ privacy, advances equity and civil rights, stands up for consumers and workers, promotes innovation and competition.”
However, executive action can only go so far. In announcing the EO, President Biden explicitly called on Congress to pass AI legislation. Senate Majority Leader Schumer has crafted a framework to guide that process. We are encouraged by several legislative proposals on the table, but we need to get this right.
We cannot be bogged down by fearmongering of an “AI apocalypse,” when there are real harms to competition and consumers that are playing out as we speak. As Vice President Harris stated recently, the harms are not only hypothetical or occurring in a far-off future; they are happening right now, to real people. How Congress handles this moment will set the tone for many years in the future. We cannot let Big Tech alone write the rules that govern itself and AI.
This memo serves to highlight the growing public support for comprehensive oversight of AI tech as Congress considers impending legislation.
When we talk about AI, we need to demystify the problem. “Artificial Intelligence” is an umbrella term encompassing the industry areas below, which deserve closer scrutiny and attention. These sectors are:
- AI Applications (AI Apps): These are end-user products that include everything from virtual assistants to autonomous vehicles.
- Large Language Models: Like ChatGPT, these models are the cognitive engines that understand and generate human-like text. They form the basis for many AI Apps.
- Cloud Computing: They serve as the foundation of AI services scalability.
- AI Chips: Specialized chips like GPUs and TPUs power data centers and form the foundation of cloud computing.
Polling makes it crystal clear: The American people want Congress to regulate AI technology and to rein in the Big Tech companies who want to monopolize it.
Polling Summary
Below are some top findings. At the end of this document, we provide a full breakdown of the various polls conducted since over the last 12 months.
Regulating AI
-
85% of respondents support a nationwide effort across government, industry, and academia to make AI safe and secure. (Source: MITRE-Harris Survey of AI Trends, July 2023, released September 2023).
-
A 62% majority of Americans across party lines support having a federal agency to regulate AI. (Source: Vanderbilt Policy Accelerator poll, September 2023).
-
A plurality of voters want Congress to take the lead on regulating AI: Respondents agreed that Congress should be the “driving force behind AI policies and regulations because it represents the public interest,” as opposed to “being driven solely by the profit motives of technology companies.” (Source: Tech Oversight Project, May 2023).
Regulating Big Tech
-
76% expressed concern over AI leading to a concentration of power in technology companies. (Source: YouGov, conducted August 21 – August 27, 2023).
-
A 75% majority of Americans across party lines think big tech companies have too much power over AI. (Source: Vanderbilt Policy Accelerator poll, September 2023).
-
A 68% majority of Americans across party lines would support proposals to break up the big AI companies. (Source: Vanderbilt Policy Accelerator poll, September 2023).
Regulating disinformation, propaganda, and deepfakes
-
85% expressed concern over AI leading to the spread of misleading video and audio deep fakes. (Source: YouGov, conducted August 21 – August 27, 2023).
-
78% of respondents expressed concern over AI leading to the spread of political propaganda. (Source: YouGov, conducted August 21 – August 27, 2023).
-
53% of respondents believe AI misinformation will impact the winner of the 2024 election. (Source: Axios, August 2023 – link to article for information on crosstabs).
See additional polling in the Appendix following this memo.
Conclusion
The growth of AI technology is already causing harm with, to name a few, predatory scams, election interference, and discrimination in housing, criminal justice, and health care. Whereas the EO could have, we argue, gone further in its protection of competition in the new market, we are encouraged by President Biden’s Executive Order and by the framework put forward by Senate Leader Schumer. However, Congress needs to pass legislation that protects people and makes sure the next generation of the technology isn’t gate-kept by some of the largest companies in the world.
The American people are clear in every poll: The time for regulating Big Tech and AI technology is now. Americans across party lines demand urgent government action to curb the unchecked influence of Big Tech.
The President has made his move. It’s now time for Congress to act.