top of page

RESEARCH REPORTS

FACEBOOK FALSEHOODS: EVALUATING DISINFORMATION RESILIENCY IN EASTERN EUROPE

August 14, 2022

During the North American Treaty Organization (NATO) Summit in June 2022, policymakers endorsed a Strategic Concept explicitly acknowledging the threat of disinformation campaigns and the urgency of preparing for such attacks in the future. To successfully fortify populations targeted by disinformation, NATO must first understand local contexts and then prescribe solutions. For example, which populations are most vulnerable to disinformation, and which are most resilient?

Screen Shot 2023-02-08 at 4.13.09 PM.png
Search
Screen Shot 2023-02-08 at 4.13.19 PM.png

EVALUATING IDENTITY BIAS IN GPT-3 GOOGLE SEARCH AUTOCOMLPLETION

January 25, 2022

I n late 2021, the technology company OpenAI released an open-access version of GPT-3, a cutting-edge artificial intelligence (AI) system that generates text in any form—including prose, poetry, and dialogue—based on a prompt given by the user. GTP-3’s text generation serves a wide number of uses in the commercial technological space.

PUBLICATIONS

EVALUATING GERMAN MEDIA’S COVERAGE OF THE RUSSIA-UKRAINE WAR

63bd76b63433c93aecd279de_christian-lue-kBk9ysFIp44-unsplash.jpeg

January 13, 2023

The student-led think tank DisinfoLab analyzed three German news outlets since the Russian invasion of Ukraine on February 24, 2022. While German mass media has generally covered the war in a manner that is sympathetic to Ukraine, tonality changed when considering the economic impacts of the war.

LE MULTICULTURALISME FAUX: THE REALITY OF BEING MUSLIM IN FRANCE

63bc6631590618337b03fbdf_khamkeo-vilaysing-SIrfBz43yYs-unsplash.jpeg

January 12, 2023

Increasing Islamophobia in France can be seen in the rhetoric of far-right leader Marine Le Pen. DisinfoLab found that Le Pen’s Islamophobic tweets increased in intensity between her 2017 and 2022 presidential campaigns, in what was likely a bid to cash in on nativism for political support.

SMALL TOWN, BIG MONEY

63bc5f267d4d44990e4994c9_element5-digital-T9CXBZLUvic-unsplash.jpeg

January 10, 2023

Over the past two years, school boards across the U.S. have emerged as focal points for heated national debates. Analysts from Disinfolab have found that national organizations have exerted influence in local school board elections in two ways: through both financial and non-financial means.

MELONI’S ELECTION ILLUSTRATES CHINA’S DWINDLING DISINFORMATION EFFICACY

WHY LATIN AMERICA IS SUSCEPTIBLE TO RUSSIAN WAR DISINFO

63bc569bf665a57c67a7e617_italy-gc1755fed1_1280.jpeg

January 10, 2023

On September 25th, 2022 Italy elected Giorgia Meloni. By analyzing Meloni’s Facebook posts and public statements, DisinfoLab, a student-led think tank at the College of William & Mary, concluded that her Sino-skeptical campaign strategy was instrumental in facilitating her election.

Screenshot 2022-09-02 091709.png

September 2, 2022

Interest groups and legislators alike have become increasingly concerned with the lack of Russian disinformation regulation in the Spanish-language.  RT en Español’s likes and comments surged on the Facebook platform. 

AI TRAINING DATA CONTRIBUTES TO ITS BIAS

opedpic2.jpeg

January 28, 2022

In an article explaining the ramifications of their report on identity bias in GPT-3 and Google Search, we break down how the sources of training data that make GPT-3 such a complex and accomplished AI also embed bias deep within its DNA.

NEW DESIGNS TO FIX THE SEARCH ENGINE BIAS PROBLEM

HOW AI BIAS IN SEARCH ENGINES CONTRIBUTES TO DISINFORMATION

RESISTING RUSSIAN AND CHINESE DISINFORMATION CAMPAIGNS

oped3pic.jpeg

January 29, 2022

In a follow-up to DisinfoLab's study comparing bias embedded in Google Search to GPT-3, we discuss the implications of these findings on the future of search engines and what it means for the spread of discrimination.

oped1pic.png

January 27, 2022

DisinfoLab conducted a World in 2050-commissioned study comparing bias embedded in Google Search to GPT-3. In this article, we discuss the implications of their findings on the future of search engines and what it means for the spread of disinformation.

Untitled design-3.png

September 13, 2022

Russia and China are actively creating and disseminating disinformation narratives in Ukraine and Taiwan to pursue their policy agendas and expand their spheres of influence. Both targeted nations have demonstrated remarkable resilience against Russian and Chinese disinformation campaigns.

INTERVIEW: DR. JOSHUA TUCKER ON BOTS AND REGULATING SOCIAL MEDIA

USING NEUTRAL NETWORKS TO PREDICT THE SPREAD OF TWEETS CONTAINING DISINFORMATION

WHAT WE CAN LEARN FROM QANON

Untitled design-3.png

August 23, 2021

At DisinfoLab, we track the activity of bot accounts on social media and seek to understand how narratives promoted by these accounts develop over time. In July, DisinfoLab sat down with Professor Joshua Tucker to discuss the relationship between social media and politics, and how bot accounts can play a role in shaping political narratives.

hoaxy_model_results.png

August 23, 2021

In Summer 2021, DisinfoLab established a Technical Analyst team to grow the research capacity of our lab. Our Technical Analysts build tools to collect large swaths of data for analysis and employ artificial intelligence to identify salient trends in disinformation.

Untitled design-3.png

August 13, 2021

If someone you loved started going around saying that the U.S. government is run by Satan-worshipping pedophiles, you might be confused. How could they believe something like that? Many Americans have faced this very question over the past four years as the QAnon conspiracy theory has gained popularity.

THE ROLE OF TWITTER BOTS IN THE BIG LIE

Untitled design-3.png

August 23, 2021

On January 6th, 2021, domestic terrorists stormed the U.S. Capitol in an attempt to stop Congress from certifying the results of the 2020 presidential election. The events of January 6th are a chilling testament to the dangers of disinformation.

BOT TRACKING: HOW DISINFOLAB MONITORS INAUTHENTIC ACTIVITY

Untitled design-3.png

August 23, 2021

DisinfoLab is a student-led research lab at the College of William & Mary’s Global Research Institute dedicated to informing American citizens about the threat of disinformation on social media platforms. Disinformation is a major threat to American citizens and democracy.

OPENAI’S “FIX” FOR GPT-3 PROVES PROBLEMATIC

6229259a0b20b087dee4327d_philipp-katzenberger-iIJrUoeRoCQ-unsplash.jpg

March 10, 2022

OpenAI in January upgraded its GPT-3 system to mitigate bias in the autocomplete language model. While this was somewhat successful, there were concerning increases in bias about gender and sexuality.

WHAT FACEBOOK COMMENTS SAY ABOUT DISINFORMATION RESILIENCE (I/IV)

62f939fa44a9e5d3c9b7b4eb_AdobeStock_450972857.jpeg

August 17, 2022

Eastern Europe is a hotbed for foreign and state-sponsored disinformation about ongoing global crises like Russia’s invasion of Ukraine and the COVID-19 pandemic. This study compared disinformation resiliency among Hungarian, Polish, and Russian speakers on Facebook by analyzing how these users responded to posts containing disinformation—ranging from agreement to rejection.

POLITICAL AND HISTORICAL FACTORS OF DISINFORMATION RESILIENCE (II/IV)

62fe286acf6c3785c46d5ef1_AdobeStock_496841688.jpeg

August 18, 2022

Disinformation about COVID-19 and Russia’s invasion of Ukraine has been rampant for the past year, but, different narratives have gained traction with different groups. The findings of the report highlight the important role of history and politics in determining which disinformation narratives will likely impact a country’s population. Therefore, when crafting solutions to combat mis/disinformation, a “one-size-fits-all” approach is bound to leave some countries more vulnerable than others.

LARGE LANGUAGE MODEL AI CAN LEAD EFFORTS TO COMBAT DISINFORMATION (III/IV)

62fbd3d48984e5834e5a7625_blurrystock-HIbAmybJHVs-unsplash.jpg

August 19, 2022

Large language models, or artificial intelligence (AI) models trained on large swaths of text, could one day expedite online disinformation research.

DISINFORMATION RESEARCH IS ABOUT TO GET HARDER (IV/IV)

62fe520cd966b057f7522f3b_shahadat-rahman-BfrQnKBulYQ-unsplash.jpg

August 18, 2022

Disinformation researchers face a common problem: collecting data. With Meta’s plan to shut down CrowdTangle in response to public relations issues, online disinformation research is poised to get harder and more legally dubious.

IMPACT OF CLIMATE DISINFORMATION: PUBLIC OPINION & LEGISLATION
(III/III)

September 6, 2022

The following publication is the final installment of our three-part series on climate change disinformation. These papers analyze the origins of climate disinformation, expose the international and domestic sources of climate disinformation, and calculate climate disinformation’s impact.

li-an-lim-ycW4YxhrWHM-unsplash.jpg

PAST LONG-FORM PUBLICATIONS

THE ORIGINS AND EVOLUTION OF CLIMATE DISINFORMATION (I/III)

August 6, 2021

The following publication is the first of a three-part series on climate change disinformation. These papers analyze the origins of climate disinformation, expose the international and domestic sources of climate disinformation, calculate climate disinformation’s impact, and brainstorm future policy recommendations.

Picture1.png
Picture2.png

SOURCES OF CLIMATE DISINFORMATION (II/III)

August 26, 2021

The following publication is the second of a three-part series on climate change disinformation. These papers analyze the origins of climate disinformation, expose the international and domestic sources of climate disinformation, calculate the impact of climate disinformation, and brainstorm future policy recommendations.

bottom of page