Generative AI (GenAI) is revolutionizing how individuals access and process information, through chat bots, search engines, email services, word processors, and web search tools, and more.1 Marc Andreessen has said AI “is highly likely to be the control layer for everything in the world” and “what AI is allowed to say/generate will be even more important—by a lot—than the fight over social media censorship.”2



But GenAI models are using misguided “misinformation” and “hate speech” policies to censor discussion, and even particular views, on controversial but important topics like COVID-19, women’s sports, and abortion.3 These ill-defined policies allow companies to censor for vague and subjective reasons and then hide behind the policies to avoid accountability. The Viewpoint Diversity Score Business Index4 found that these terms are already pervasive in “Acceptable Use Policies” at Alphabet] and virtually every other major tech company.



But the loudest voices are calling for more, not less, censorship. The World Economic Forum ranked AI-generated “misinformation and disinformation” as the #1 global risk, above inflation, climate change, and societal polarization.5 Alphabet shareholders asked for a report just last year on the same.6 And laws in the EU and 21 countries mandate or pressure digital platforms to deploy AI chatbots to remove disfavored political, social, and religious speech.7



The above policies reach well beyond legitimate harms and target sweeping amounts of protected expression. They also ignore the critical role of free speech and freedom of conscience which lie at the heart of a free society. These rights are listed first in the U.S. Bill of Rights and are protected in Articles 18 and 19 of the United Nations’ International Covenant on Civil and Political Rights. Alphabet, for its part, has committed to upholding these and other core human rights8 and to “[a]void creating or reinforcing unfair bias” in its 2018 AI Principles.9



As Time Magazine reported, “GenAI should augment, not replace, human reasoning. This critical function is hampered when guardrails designed by a small group of powerful companies refuse to generate output based on vague and unsubstantiated claims of ‘harm.’”10



The fall of the Global Alliance for Responsible Media11 and Gemini’s debacle with black founding fathers and Nazis12 show that this heavy-handed censorship is wildly unpopular. It also raises serious legal risk because of concerns about illegal discrimination.



Alphabet must address these concerns and assure its shareholders that it is protecting fundamental freedoms that will contribute to a free and fair society and serve its diverse users and customers.