Community Notes and Fact-Checking

 Community Notes are better than you may think

    I think that everyone has heard of Zuckerberg's plans to discontinue fact-checking cooperations in favour of Twitter-like community notes to delegate the service to the users of Meta services[1]. Many have complained that he thus gave away responsibilities for the paddling of misinformation and disinformation, more specifically its combatting, to the users, implicitly conceding to incoming POTUS Donald J. Trump. To be fair, on that latter point, I agree to some degree, although I also believe that he tries to play it cool with him as he may remember that he threatened to imprison him on the basis that is either pure retaliation or an unofficial lèse-majeste[2]. Whatever the reason may be--perhaps Ms. Telnaes was right and gradually more wealthy entrepreneurs and businesspeople kneel before the authoritarian president not only to be left alone but to appeal to him for personal or corporate benefits[3]--, but what I want to focus on is to take up the cudgels on behalf of community notes as a means of fact-checking. They are much better than their reputation gives them credit for.

     On Twitter, they have been around for quite some time now, a “relic” from the time before Tesla CEO Elon Musk took over the reins from Jack Dorsey, and many still-existing users who have not gulped down Musk's Kool-Aid showed genuine surprise that he didn't yet discontinue them on his platform either. (A question that was also asked before the European Union passed the Digital Services Act, abbr. DSA[4]) Whatever the reason may be, it is good that they received implicit graces from the CTO, formerly CEO. One cannot expect that any such nudging off the path of fake news and conspiracy theories--those who are already deep into the respective rabbit hole will not be lured outside by either fact-checks published with resources they already taint as “propaganda organs of the established powers”--would either immediately or gradually “purify” them again[5]. The whole concept of believing in the “right” opinion could only be successful through indoctrination. But this is not what we can wish to achieve. Instead, we should focus on the implementation of critical-thinking skills, although neither community notes, nor fact-checking work could do. That's where educational systems must pick them up, at best once they start going into school[6]. After reaching adulthood, it required personal interests which the believers of conspiracy theories preferred to abstain from. Igorance is bliss, after all...

    The question is: Are community notes any good? Many critics have spoken out against them, calling them ineffective at combatting disinfromation[7]. But what does research say? One study draws a more favourable picture of their performance, proving that at best, authors of community notes are quicker to the scene than the employed experts professing in fact-checking[8]. To me as someone who cherishes factuality and context to fill informational holes, this makes sense, also helping provide complete frames. It secondly helps to provide opposition to the doomsayers who claim that disinformation reigned supreme and beyond bot networks operated from afar in Petersburgian offices[9]. 

    Another factor that should be taken into consideration is the engagement factor in community notes: Populists around the Western world laud “direct democracy” as they understand it because power were handed back to the public; the reason why they call it the true Democracy, instead of the parliamentary system dominant in Europe and the Americas. Under the existing conditions, it were unsuitable as the masses cannot effectively be dleegated questions concerning foreign affairs and national security, but here is not the place to discuss such topics. The reason I do mention it is because of the engagement it effectuated once it were installed: People had to at best inform themselves more as now, they had to fell decisions previously delegated to ministers, secretaries and their employees. The daily digests on the TV, radio or news websites (not to speak of newspapers) no longer sufficed: They had to skim through briefings, expert analyses and whatever were available in order to make a prudent choice. 

    Community notes, in a way, are similar to that: In order to fact-check posts, it didn't suffice to just write that someone were wrong and why they were wrong: You also had to provide sources that proved your point. Otherwise, you wouldn't receive the votes of support to rate your note higher[10]. People may feel a sense of gratification when their community notes receive appreciation, Or perhaps they feel as if they did their community something good, the same feeling of which honorary offices live off too[11]. 

Image by KLAU2018

    Now, I haven't made this up as a thesis of my own. Scientists have examined this theory and have come to the conclusion that there must indeed be such a driver that keeps the community notes alive as a viable means of fact-checking posts on Twitter, and we can assume that there will be a similar noticeable effect on Facebook, given that its users don't necessarily differ all that much. Furthermore, because of the opportunity to engage with the fact-checking process also improved the overall trust in their credibility, although the sources relied upon in the fact-checking process didn't differ from those applied by “anonymous” fact-checkers employed by the social media company[12]. The visibility may be the key to the increased trust, as opposed to “anonymous” newspapers and their staff, although there is good reason why some of them prefer to stay hidden. Harassment increases still, peaking in the murder and sudden disappearances[13]. 

    As I have said beforehand, what fact-checking and community notes cannot do is to slow down the engagement with fake news andn conspiracy theories, inter alia[14]. But there can be no other option than to prescribe what we too have mentioned beforehand: That people must be educated in thinking critically[5] in order to work with the informational deluge that presents itself to them daily. Whatever attempt at regulating gatekeepers on social medica were tried has to fail inevitably; not because of the heads' disinterest in complying with those enquiries.The only measurement that could help would be an examination of each and every post before it were finally released to the public, which would break down the entire environment of a social medium as a whole. And on a more personal note, fact-checking of reported content is a job that leaves a human being miserable sooner or later, as earlier reports showed[15]. Perhaps AI tools will someday be able to properly and fairly moderate content, either at request or autonomously, but at the moment, we are not at such a point that this process could be fully moderated without human supervision to process motions of appeal by affected users[16]. 
    What does this leave us at? In my opinion, community notes are better than their reputation, they have got a true potential to replace fact-checking as an institutionalised service. So far, we shouldn't call them out as a failed attempt to appease both Trum pand the more serious legislators. Studies have shown that they can do the trick and even achieve a surplus in confidence amonst the more sceptical individuals, i.e. those who need to be reached out to. The only reason that I could think of why some people may be suspicious is because they view fact-checking as a whole as a panacea to the ailments that occur to us every day on the internet. Expectations, though, must be lowered because in the end, they only provide context and corrections/rectifications to statements iterated by others. What the reader makes out of this is up to them; it does not destine them to radically overhaul their opinions. They may still discard them as nonsense pushed by “established media” &c. 

I, for myself, remain hopeful. 

---

Footnotes

[1] https://about.fb.com/news/2025/01/meta-more-speech-fewer-mistakes/

[2] https://apnews.com/article/trump-book-zuckerberg-prison-election-donations-putin-b717f6248311b3ed002d872ec97d59d9

[3] https://anntelnaes.substack.com/p/why-im-quitting-the-washington-post 

[4] Peukert, A. (2023). Who Decides What Counts as Disinformation in the EU?. Verfassungsblog. Link: https://verfassungsblog.de/who-decides-what-counts-as-disinformation-in-the-eu/

[5] Lewandowsky, S., Ecker, U. K. H., Seifert, C. M., Schwarz, N., & Cook, J. (2012). Misinformation and Its Correction: Continued Influence and Successful Debiasing. Psychological Science in the Public Interest, 13(3), 106–131. https://doi.org/10.1177/1529100612451018
It would also require one to find out why specific groups of people begin to dive into conspiracy-theory rabbit holes: 
Bowes, S. M., Costello, T. H., & Tasimi, A. (2023). The conspiratorial mind: A meta-analytic review of motivational and personological correlates. Psychological Bulletin. DOI: https://doi.org/10.1037/bul0000392.supp 

[6] Lantian, A., Bagneux, V., Delouvée, S., & Gauvrit, N. (2021). Maybe a free thinker but not a critical one: High conspiracy belief is associated with low critical thinking ability. Applied Cognitive Psychology, 35(3), 674-684. DOI: https://doi.org/10.1002/acp.3790
The same conclusion was also drawn years later with reference to media literacy as a means to evaluate what one reads on the internet:
Yeoman, F., & Morris, K. (2023). The Place of Media Organisations in the Drive for Post-pandemic News Literacy. Journalism Practice, 18(1), 158–180. https://doi.org/10.1080/17512786.2023.2169186

[7] https://www.washingtonpost.com/technology/2024/10/30/elon-musk-x-fact-check-community-notes-misinformation/

[8]   Luca Righes, Mohammed Saeed, Gianluca Demartini, Paolo Papotti. The Community Notes Observatory: Can Crowdsourced Fact-Checking be Trusted in Practice?. WWW 2023, Companion  Proceedings of the ACM Web Conference„ ACM, Apr 2023, Austin, United States. pp.172-175,  https://doi.org/10.1145/3543873.3587340 . hal-04087037

[9] It has been proven anyway that Russian bots had less of an impact on the 2016 general election than it was believed shortly after its conclusion, when Donald J. Trump won the presidential election the first time:
Eady, G., Paskhalis, T., Zilinsky, J. et al. Exposure to the Russian Internet Research Agency foreign influence campaign on Twitter in the 2016 US election and its relationship to attitudes and voting behavior. Nat Commun 14, 62 (2023). https://doi.org/10.1038/s41467-022-35576-9
    As for the 2024 General Election, it is yet too early to evaluate the impact it had on the election, although it should not be as much as the tides were already turning against the Democrats due to their infighting because of Biden's rather weak outlook, given his age and physical, as well as mental, deterioration. But these are just my unsubstantiated two cents, I still await the first analyses.
    Pundits in online political zines have already wondered, therefore, whether Americans received the government they wanted, plus the ruler they craved for: “a cold-hearted Republican to lower taxes, brutalize criminals, and rule you like a king”, as Sideshow Bob said in The Simpsons (S06E05): https://www.thebulwark.com/p/how-authoritarian-are-americans-trump-surveys-autocracy

[10] https://communitynotes.x.com/guide/en/about/introduction

[11] Hübler, O. (2024). Donations, volunteering, and life satisfaction in Germany. Economics Bulletin. Link: https://www.accessecon.com/Pubs/EB/2023/Volume43/EB-23-V43-I4-P164.pdf (PDF, 371 KB)

[12]  Chiara Patricia Drolsbach, Kirill Solovev, Nicolas Pröllochs, Community notes increase trust in fact-checking on social media, PNAS Nexus, Volume 3, Issue 7, July 2024, pgae217, https://doi.org/10.1093/pnasnexus/pgae217

[13] Orgeret, K. S., Westlund, O., & Krøvel, R. (2024). Journalism and safety: Digital threats, professional fragilities, and safety cultures. Routledge. DOI: https://doi.org/10.4324/9781003544517
    As you could expect, the RSF's report of last year didn't strike an optimistic tone concerning the global state of journalists' safety in duty: https://rsf.org/en/rsf-s-2024-round-journalism-suffers-exorbitant-human-cost-due-conflicts-and-repressive-regimes

[14] Yuwei Chuai, Haoye Tian, Nicolas Pröllochs, and Gabriele Lenzini. 2024. Did the Roll-Out of Community Notes Reduce Engagement With Misinformation on X/Twitter?. Proc. ACM Hum.-Comput. Interact. 8, CSCW2,  Article 428 (November 2024), 52 pages. https://doi.org/10.1145/3686967

[15] https://www.theverge.com/2019/2/25/18229714/cognizant-facebook-content-moderator-interviews-trauma-working-conditions-arizona

[16] Udupa, S., Maronikolakis, A., & Wisiorek, A. (2023). Ethical scaling for content moderation: Extreme speech and the (in)significance  of artificial intelligence. Big Data & Society, 10(1). https://doi.org/10.1177/20539517231172424

Nahmias, Y., & Perel, M. (2021). The oversight of content moderation by AI: impact assessments and their limitations. Harv. J. on Legis., 58, 145. Perm. Link: https://heinonline.org/HOL/LandingPage?handle=hein.journals/hjl58&div=7&id=&page=