Sen. Cruz Sounds Alarm Over Industry Role in AI Czar Harris’s Censorship Agenda

Letter Questions RAND's Part in AI EO Drafting, Citing Past Censorship Work and Ties to Major AI Companies Funding Harris's Campaign

WASHINGTON, D.C. - U.S. Senate Commerce Committee Ranking Member Ted Cruz (R-Texas) is questioning the Biden-Harris administration's collaboration with the RAND Corporation, a think tank with Big Tech ties, citing RAND's previous involvement in both the drafting of the Biden-Harris Artificial Intelligence (AI) Executive Order and efforts by the government to censor online speech. In the letter to RAND's president and CEO, Sen. Cruz highlights that RAND has received significant funding from tech billionaires who support the controversial "effective altruism" movement. The letter also details how RAND and left-wing groups are embedding AI staffers in federal agencies. Additionally, some of these Big Tech leaders are financially backing Vice President Harris's presidential campaign, raising alarms about potential conflicts of interest in AI regulation.

The letter details RAND's past involvement in developing a controversial left-wing censorship model within the government "to track the spread of false and misleading information" via Meta's new virtual reality system.

The concerns about the Biden-Harris administration's AI censorship efforts are especially significant in light of recent news that Meta CEO Mark Zuckerberg admitted to working with the Biden-Harris administration to censor COVID-19 information and acknowledged that Facebook wrongly suppressed the New York Post's story about Hunter Biden's laptop and Biden family corruption.

Ranking Member Ted Cruz writes RAND President and CEO Jason Matheny:

"It is no secret that under the guise of protecting the public from various, subjective and amorphous harms, the Biden-Harris administration engaged in "a far-reaching and widespread censorship campaign" against Americans who engaged in so-called "misinformation" on social media platforms."

"Of course, the administration could have more easily surveilled Americans had the government acted to regulate the social media industry in its infancy. Recognizing Artificial Intelligence (AI) regulation as a new opportunity to spy on Americans and restrict speech, last year the Biden-Harris administration launched a campaign to "protect" the public from undefined "harms" that AI could cause."

"As the first order of business, the White House announced Vice President Harris-who described AI as "kind of a fancy thing" that is "two letters'-would become the 'AI Czar' responsible for leading this new regulatory initiative."

In the letter, Sen. Cruz expresses concerns about AI Czar Harris's efforts to control AI development and influence the content Americans see online. He specifically requests information on RAND's outsized role in drafting the AI Executive Order (EO 14110), given that a number of organizations that worked with the administration to censor Americans on social media were involved in the AI Executive Order effort:

"Then the administration issued Executive Order 14110 on the 'Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence' (AI Executive Order), which sets forth an audacious plan for the government to control AI development and determine, at the technology's outset, what content Americans can see and share."

Moreover, the individuals and organizations involved in shaping the administration's new AI regulations are the very same ones who helped the administration censor Americans' social media posts."

... I write seeking information regarding RAND's role in drafting the AI Executive Order and its previous involvement in the administration's censorship agenda.

Shortly after taking office, the Biden-Harris administration announced it would address what it called a 'crisis of disinformation and misinformation.' The administration has addressed this supposed 'crisis' by censoring Americans' speech that it does not like, from coercing social media companies to adopt strict content moderation policies and removing posts about the COVID-19 lab leak theory, to establishing a misinformation team within the Cybersecurity and Infrastructure Security Agency that 'counters all types of disinformation, to be responsive to current events.'

With the emergence of AI as the latest technology that will have an enormous impact on speech, including the spread of opinions and content disfavored by the censorious Left, the administration is seeking to restrict AI development through regulatory fiat. And in doing so, it is citing the same amorphous 'harms' as it did to justify social media censorship...

Sen. Cruz argues the AI Executive Order aims to control speech and stifle innovation and competition by requiring companies to submit detailed reports to the government certifying that they conform to woke AI "safety" standards, while simultaneously entrenching the most powerful industry players.

"Last fall, the White House issued its sweeping AI Executive Order, section 4 of which details the administration's plan to control speech so AI is developed 'safely and responsibly...'

"...The AI Executive Order also empowers federal bureaucrats to monitor the activities of all companies developing, or even considering developing, an AI model. Specifically, section 4 requires that wherever a company contemplates or develops a 'dual-use foundation model,' it must submit to the government records of all AI training, development, and production, and (most alarmingly) 'safety evaluation reports,' describing how well the AI conforms with NIST's woke AI 'safety' standards.

Altogether, the AI Executive Order sets the path for the federal government to preemptively discourage AI innovation and protect powerful and established AI companies.

In the letter, Sen. Cruz questions RAND's ties to established AI companies and their funding of Harris's presidential campaign.

"RAND apparently had an outsized role in drafting the AI Executive Order-and Section 4, in particular. You openly admitted that RAND assisted the White House in drafting the AI Executive Order days before the President signed it..."

"Indeed, the AI Executive Order captures all six of the policy recommendations a senior RAND official provided in congressional testimony last year."

"RAND has both personal and financial connections to incumbent AI companies like Anthropic and OpenAI. For instance, you are one of five members on Anthropic's 'Long-Term Benefit Trust,' which oversees the AI company. And Tasha McCauley-an adjunct senior management scientist at RAND and effective altruism advocate-just left OpenAI's board in March."

"RAND is also connected with Open Philanthropy, the charitable arm of these AI companies. Last year, Open Philanthropy gave RAND $16 million to advance its 'work on potential risks from advanced artificial intelligence...'"

"Given AI Czar Harris's outspokenness on the need for government to protect against AI's so-called dangers, like disinformation, it is unsurprising that the leaders of these organizations and companies are funding Harris' presidential campaign."

Sen. Cruz highlights RAND's past support for social media censorship and government surveillance. The letter cites contracts awarded to RAND by the Department of Homeland Security (DHS) for so-called mis?, dis?, and mal-information (MDM) research notes that watchdog groups have expressed concerns about these efforts leading to the censorship of certain viewpoints.

Before drafting the AI Executive Order, RAND supported the administration's social media censorship work. For example, a 2023 report from RAND's Homeland Security Research Division, which operates the Department of Homeland Security's (DHS) Homeland Security Operational Analysis Center, calls for the government to spy within the metaverse 'to track the spread of false and misleading information,' while warning that it may be 'difficult' to do so.

Indeed, DHS gave RAND taxpayer money for specific MDM-related research. In September 2021, DHS awarded RAND a contract for research to help DHS understand which component missions were vulnerable to MDM information.^[ And in June 2022, DHS awarded RAND a sole-source contract to 'develop[] an MDM research agenda to support counter-MDM programming.'

Watchdog groups have raised concerns that awards like these have resulted in the censorship of conservatives and Christians.

The letter emphasizes that RAND is aligning its research with the Biden-Harris administration's AI Executive Order and is pushing for more permanent legislative measures, including content moderation requirements and holding generative AI developers accountable for material produced by users of their platforms.

"RAND appears to be shifting its research to support the Biden-Harris administration's AI Executive Order. It recently published a paper on generative AI..."

"...The paper warns 'generative AI opens up even newer avenues of both malicious use and unwitting misuse of information, which can lead to . . . the spread of both misinformation and disinformation."

"'It further cautions that the AI Executive Order 'could be overturned by a new president' and therefore recommends 'the U.S. Congress to consider more permanent legislative approaches.'"

"One approach the federal government should consider, RAND says, is 'set[ting] requirements for basic content moderation regimes.' In setting up these regimes, RAND proposes regulations 'focus on the ways in which AI platforms perform moderation, such as by stipulating requirements for training data or requiring input/output filters.' In the alternative, RAND suggests making AI developers liable for their content."

RAND's push for permanent legislation is significant given that the Senate Commerce Committee recently advanced six AI bills, including the Future of Artificial Intelligence Act, that effectively codify the Biden-Harris AI Executive Order. If enacted, this legislation would consolidate AI development among a handful of America's largest tech companies and facilitate government-Big Tech collusion far beyond what was pursued during the COVID pandemic and the 2020 election.

The letter requests information about RAND personnel who worked for or advised the federal government on the AI Executive Order before its publication. In addition, the letter requests documents related to misinformation, disinformation, or AI research conducted for the federal government. Sen. Cruz encourages those with knowledge of these matters to contact his office directly at [email protected].

Read Full Letter HERE.

###

https://www.commerce.senate.gov/2024/9/sen-cruz-sounds-alarm-over-industry-role-in-ai-czar-harris-s-censorship-agenda