Govt Funds AI Tools for Censorship

According to a government committee report the National Science Foundation is funding the development of automated tools to censor online speech “at scale.” This is the conclusion reached in the 80 page Interim staff report of the committee on the judiciary, and the select subcommittee on the weaponization of the federal government, US House of Representatives, February 5, 2024, titled “The weaponization of the National Science Foundation: How NSF is funding the development of automated tools to censor online speech at scale and trying to cover up its actions.”(1) 

The National Science Foundation is the civilian arm of DARPA: the Defense Advanced Research Projects Agency. DARPA is the national security state cloaked in civilian dress. It is a 10 billion dollar source for university project funding. 

This slide tells the story.(2) The government awards contracts to universities and non-profit corporations, their research is turned into censorship and propaganda software tools, these tools are deployed and used to collate and tag posts, which are then removed from social media platforms. 

The Judiciary Report warns that, 

Although any violation of the first amendment is alarming, the EIP’s [Election Integrity Partnership] effort led to only thousands of Americans posts being targeted; new technologies could enable a much smaller team to accomplish the same task for millions of posts, if not entire narratives.(3) 

The Report quotes Marc Andreesen, co-founder of Netscape, 

AI is highly likely to be the controller for everything in the world. How it is allowed to operate is going to matter perhaps more than anything else has ever mattered. You should be aware of how a small and isolated coterie of partisan social engineers are trying to determine that right now, under cover of the age-old claim that they are protecting you.(4) 

Supposedly those who would deliver us from “disinformation” are protecting us. One example of this “protection” in operation is called, 

…WiseDex harnesses the wisdom of crowds as well as AI techniques to select keywords for each claim and provide other information in the claim profile… WiseDex enables fast, comprehensive and consistent enforcement around the world, so that harmful misinformation stops reaching big audiences.(5) 

The tool is described in notes accompanying a slide deck as a means of “externalizing the difficult responsibility of censorship.”(6) According to the Judiciary Report, the University of Michigan received a $750,000 grant from the National Science Foundation(7) to build software able to empower AI scan-and-ban artificial intelligence machine learning-powered scripts. Mike Benz stated that the tool is used to categorize claims made online about vaccines, masks and mandates, and has been used in the creation of a database and the creation of scripts to run on the algorithms at Facebook, Twitter and TikTok so that those who dared speak on these topics were tagged and intercepted at the level of the algorithm.(8) 

Externalizing censorship is necessary because, as Renee DiResta at the Stanford Internet Observatory wrote in another set of slide notes, there were “unclear legal authorities including very real first amendment questions.”(9) That is an understatement. The First Amendment to the Constitution prohibits “abridging the freedom of speech.” 

The people being paid to develop these tools say the opposite. University of Michigan slide notes state, “There is widespread agreement misinformation undermines liberal democracy by eroding the commonly accepted facts that allow society to reach consensus…. Our work reestablishes bipartisan consensus on truth, and therefore enables platforms to curtail the spread of misinformation, saving modern democracy.”(10) What does today’s government apparatchiks deem to be misinformation or disinformation? Consider this clip from a government online training event: 

In the playbook we give a definition for disinformation as false information that is deliberately spread with malicious intent, and misinformation that is false information that’s spread mistakenly. Our approach in giving recommendations to counter mis or disinformation is to think of them as one concept, not because they’re the same but because when an election official, or a voter, or others might encounter mis or disinformation, they will not be able to tell necessarily the intent, or whether it’s disinformation or misinformation incident.(11) 

Those definitions propose that someone gets to abridge free speech by deciding for us what is and is not false information. Social media companies can act on content based upon information provided via authoritative external entities, thus abridging free speech. 

The situation may actually be yet more serious. In a 2024 interview, Mike Benz said, 

AI censorship is not some future threat; it’s already here. It’s already been here for about six years. It was the great technological game changer that allowed the modern era of Internet censorship to happen after 2016. If political elites wanted to kill a narrative online that was trending or a political movement that they opposed, they would have to hire armies of manual censors… 

Benz says that even after hiring tens of thousands of employees to tag social media posts for censorship, the manual method was insufficient, which led to the development of the new weapons to automate censorship. 

This technology was initially created by DARPA using a technique called natural language processing, a machine-learning technique that takes every word you say, every sentence you write, and every paragraph you tweet on Twitter. All is analyzed in a model ingesting what you say, as well as how it spreads to the other people in your network. They create vast topographical network maps of how a narrative is going viral... They can create this taxonomy of any narrative in the world—vaccine skeptics, mail-in ballot skeptics, climate skeptics, abortion, energy, migration. This was initially constructed by the pentagon with funding from the state department in order to deal with the threat of Isis recruiting on Facebook and Twitter in 2014… All of this goes into a back-end text-to-speech transliteration that’s used for closed-captioning, which means all that is translated into words. 

According to Benz, now, 

“…it creates for the national security state a real-time heat map of every political movement and of every trending or emerging narrative that’s going on in the world. So it was initially created for those national security purposes, but then after Brexit in the UK and the 2016 election here in the US, when they created this Russiagate predicate all of that came home, and they turned these weapons of mass-deletion to be able to scan and ban any narrative at scale from a foreign national security predicate  into a domestic democracy predicate and unleashed it on everyone.”(12) 

In 1521 in Germany, Emperor Charles the V issued the edict of Worms, a command, among other things, banning the writings of Martin Luther. It was state censorship. The Protest of the German princes in 1529 secured freedom of thought and of conscience and became the source of the name “Protestantism.”(13) The Bible in the book of Revelation foretells an authoritarian power global in scope, able to enforce its commands and doubtless its narratives (Revelation 13:14). Capability to deamplify is also ability to amplify, to increase support for an idea. A censorship machine is also a propaganda machine, able to amplify popular demand. Dissenters will be unable to buy or sell (Revelation 13:16). No government in human history has possessed the machinery to enact censorship at scale—until now.

****

Larry Kirkpatrick serves as pastor of the Muskegon and Fremont MI Seventh-day Adventist churches. His website is GreatControversy.org and YouTube channel is “Larry the guy from Michigan.” Every morning Larry publishes a new devotional video.


Notes

1. Interim staff report of the committee on the judiciary, and the select subcommittee on the weaponization of the federal government, US House of Representatives, February 5, 2024, titled “The weaponization of the National Science Foundation: How NSF is funding the development of automated tools to censor online speech at scale and trying to cover up its actions,” https://judiciary.house.gov/sites/evo-subsites/republicans-judiciary.house.gov/files/evo-media-document/NSF-Staff-Report_Appendix.pdf, accessed 2024-02-08 (hereafter, Judiciary Report).

2. Ibid., p. 2.

3. Ibid., p. 7.

4. Ibid., p. 9.

5. U.S. National Science Foundation, Track F: WiseDex // Phase I Project Video, June 17, 2022, https://www.youtube.com/watch?v=18gNRQaQtfw, accessed 2024-02-15.

6. Judiciary Report, p. 15.

7. Ibid., p. 15.

8. Patrick Henningsen Show, TNTRadioLive, “Interview: Mike Benz ‘A.I. Censorship Will Play Role in 2024 Election,” https://rumble.com/v4cc1p2-interview-mike-benz-a.i.-censorship-will-play-role-in-2024-election.html, accessed 2024-02-15.

9. Judiciary Report, p. 24.

10. Ibid., Appendix C.

11. FFOSourceClips, “DHS ‘Disinformation Training—Harvard Partners Teach Not to Distinguish Between Mis and Disinformation,” https://rumble.com/v1ha7v1-dhs-disinformation-training-harvard-partners-instruct-not-to-distinguish-be.html, accessed 2024-02-13.

12. Patrick Henningsen Show, Ibid.

13. “One of the noblest testimonies ever uttered for the Reformation was the Protest offered by the Christian princes of Germany at the Diet of Spires in 1529. The courage, faith, and firmness of those men of God gained for succeeding ages liberty of thought and of conscience. Their Protest gave to the reformed church the name of Protestant; its principles are “the very essence of Protestantism,” Ellen G. White, The Great Controversy, p. 197.