The Threat to Free Speech Posed by AI Bias Isn’t Artificial

Google headquarters in New York City, December 17, 2018 (Jeenah Moon/Reuters)

We are witnessing the censorship–industrial complex in overdrive.

Sign in here to read more.

We are witnessing the censorship–industrial complex in overdrive.

O ne of the great television villains of all time, known to Jerry Seinfeld and the world simply as “Newman,” once said, “When you control the mail, you control . . . information!”

These days, while most Americans might occasionally receive essential information in their mailboxes, the gatekeepers to information are digital, and the power to control the free flow of information rests in a dangerously small number of large companies. As our nation approaches a presidential election, citizens who take civic duty seriously are searching for information about candidates to cast an informed vote. The search results are disturbing, to say the least.

Google’s auto-complete feature for the search engine is hiding factual information about one of the biggest news events in American political history. This includes omitting keyword prompts on “assassination attempt on President Trump” from search engines. Google has admitted to the problem, which is the first step to recovery. Meta AI’s chatbot provides evasive answers to basic questions about the event and falsely flagged iconic images as “altered” on Facebook and Instagram.

We are witnessing the censorship–industrial complex in overdrive. When online search tools powered by AI-driven auto-complete algorithms cannot or will not provide factually relevant answers, it raises the specter that these companies are intentionally trying to manipulate public opinion. George Orwell coined the phrase “memory hole” for situations like these. It is as if the oligarchs of technology read Orwell’s dystopian prophecies not as cautionary tales warning against the dangers of concentrated power but as instruction manuals.

This Big Tech misinformation campaign reflects a dramatic expansion of earlier instances of AI bias, including Google Gemini’s unprompted generation of racially diverse Founding Fathers, Vikings, and a pope, as well as its inability to generate images for terms such as “white family” or “white scientist.”

Evidence presented in the recent U.S. Supreme Court case Murthy v. Missouri should remove all doubt that major corporations like Google and Facebook work with government regulators to hide information from the American people. At oral argument, the Court seemed to struggle in defining the government’s conduct as coercive. Ultimately, the Court did not address the merits of the case, instead holding that the individuals whose speech was censored lacked standing to bring the case. The FBI, emboldened by the Supreme Court, just promised to re-up its “partnership” with social-media platforms to censor “misinformation.” Whether the platforms act under coercive duress or as willing co-conspirators, the results are the same. Unfortunately, it appears that the companies that dominate this market have established manipulation as customary practice by this point. Big Tech’s smaller competitors are returning similar search results.

The American people deserve answers to what is going on, and we deserve them now. If Americans cannot even access facts from the primary source of information in the digital age, then we lack a fundamental element of a free society. Accurate information is necessary for reasoned civil discourse; it is oxygen to the body politic.

If the companies refuse to come clean, then Congress and the states must demand transparency and accountability by using every legal tool available. In the case of Google, some have suggested an antitrust action to break up the only search engine in the world whose name is used as a verb in common parlance, which would be the nuclear option. But if the Cold War taught us anything, demonstrating the will to use the most extreme option can prevent the necessity.

For years, tech giants such as Google and Meta have influenced public discourse, sometimes in the shadows. Recent examples make the threat to free speech more apparent than ever. Auto-complete and AI assistants, which too often eagerly highlight recent political-campaign activity from one political party while either suppressing or ignoring the other, are eroding the public’s ability to make informed choices — the foundation of a democratic republic.

Concentrated power is anathema to liberty. The founders of our republic crafted our Constitution with one overarching goal: to limit the potential for concentration of power in the federal government. This concentration of power in the information age is just as dangerous. If a small cadre of companies is allowed to manipulate information about a national election for the highest office in our government, we as a nation are entering uncharted territory.

Without a course correction, we run the risk that future generations may wonder what it was like to find facts, not propaganda, on the internet and cast a truly informed vote. They will have to Google it.

Lathan Watts is the vice president of public affairs for Alliance Defending Freedom (@ADFLegal). He served for three years as the Dallas coordinator for National Review Institute’s Burke to Buckley Fellowship and earned his juris doctor degree from the University of Mississippi.
You have 1 article remaining.
You have 2 articles remaining.
You have 3 articles remaining.
You have 4 articles remaining.
You have 5 articles remaining.
Exit mobile version