Sage Journals: Discover world-class research

Abstract

Public trust in the news media has eroded in the United States. This study examines how perceptions of misinformation (PMI) and disinformation (PDI) affect the consumption of traditional media, social media, and artificial intelligence (AI) news, and whether this relationship is moderated by political ideology and media trust. Findings from a pre-registered experiment (N = 637) revealed that PMI and PDI regarding traditional and social media news lowered intentions to consume news from traditional, social, and AI sources. We found no significant moderating effect of political ideology or media trust. The implications of the findings are discussed.

Keywords

news misinformation disinformation artificial intelligence traditional media social media

Trust in the news media has declined globally, with a noticeable decrease observed in the United States (Hanitzsch et al., 2018). As skepticism toward news media outlets grows, people turn toward alternative news channels for information (Tsfati & Cappella, 2003). This shift is likely influenced by increasing perceptions of misinformation (PMI) and disinformation (PDI) in mainstream news, which can impact trust in news media and further alter traditional news consumption behaviors (Hameleers, 2022). PMI can be conceptualized as negative perceptions about the unintentional lack of reliability and inaccuracy of news, whereas perceptions of disinformation reflect the idea of intentionally reporting false information to mislead (Hameleers, Brosius, & de Vreese, 2022; Pinkleton et al., 2012). A consequence of these perceptions could be lower trust in the news media (Van Aelst et al., 2017) and increased use of alternative news media (Benkler et al., 2018).

Amid declining trust, artificial intelligence (AI) offers an alternative source of information to traditional media. AI refers to technology that is equipped with capacities that are associated with intelligence, such as machine learning, natural language processing, communication, problem-solving, reasoning, planning, learning, computer vision, and speech recognition (Russell & Norvig, 2010; Rzepka & Berger, 2018). AI technologies include conversational agents like ChatGPT and Microsoft Bard, voice assistants like Siri and Alexa, and autonomous systems like Tesla’s self-driving cars. These tools are capable of understanding commands, providing recommendations, and improving their performance over time through learning.

With recent technological advancements, AI technologies are also increasingly used to access information or provide personalized news (Simon & Isaza-Ibarra, 2023). As AI becomes a more prevalent source of information, understanding how PMI and PDI affect individuals’ news media consumption patterns is crucial. Furthermore, AI may be perceived as a useful alternative channel for providing unbiased, aggregated news by the public.

Theories of selective exposure and motivated reasoning indicate people tend to select media sources that align with their pre-existing attitudes over those that are incongruent with their pre-existing attitudes (Hart et al., 2009; Kunda, 1990). Individuals may gravitate toward alternative media channels, such as AI, rather than traditional news media when they perceive that news media spread mis/disinformation. Similarly, PMI/PDI may be dependent on political ideology, where individuals may prefer media that align with their political identity (Garrett & Stroud, 2014). Recent work by Mourão and colleagues (2018) found that conservatives are less likely to trust the news compared to liberals. Partisans may be more likely to shift media sources when they perceive mis/disinformation to be present in traditional sources, particularly those that challenge existing views—likely due to the hostile media effect, which suggests that individuals find belief-challenging sources as hostile, resulting in reduced trust and use of such sources (Gunther et al., 2017).

This research involves a pre-registered experiment that examines the causal relationship between PMI/PDI and individuals’ intentions to consume traditional and social media news and adopt AI as an alternative news source. We also test whether traditional media news trust moderates the relationship between PMI/PDI and news use.

Literature Review

Understanding Perceptions of Mis/Disinformation

As trust in media has declined, individuals view traditional and social media news sources as dishonest and unreliable for obtaining information (Bachmann & Valenzuela, 2023). Trust in news media can be influenced by multiple factors, such as individuals’ pre-existing beliefs and how the information in the news is framed (e.g., negativity bias; Lindgren et al., 2024), as well as perceptions that incorrect information is being presented (Hameleers, Brosius, Marquart, et al., 2022). Mistrust in news media due to the presence of false information has been conceptualized into perceptions of mis/disinformation. Scholars have defined misinformation as inaccurate information that does not necessarily intend to deceive but can be regarded as being untruthful, as well as information that contains errors or is incorrect based on experts’ knowledge (Hameleers, Brosius, Marquart, et al., 2022; Vraga & Tully, 2021). Meanwhile, disinformation is pertinent to purposeful manipulation and fabrication, which is a more extreme form of misinformation (Hameleers, Brosius, Marquart, et al., 2022).

News consumers proactively decide which media channels they wish to rely on for obtaining information, which may be contingent on different levels of trust and skepticism toward media channels (Hameleers, Brosius, & de Vreese, 2022). It has been implied that those with greater trust in the media, in general, are more likely to expose themselves to mainstream media when it comes to news source selection and those with greater media skepticism tend to lean toward choosing non-mainstream, alternative media sources (Fletcher & Park, 2017; Rasul et al., 2025; Tsfati & Cappella, 2003). On the contrary, exposure to alternative media has been associated with exacerbating PMI/PDI (Egelhofer & Lecheler, 2019). This means that reliance on alternative media sources that attack the mainstream media could increase one’s tendency to believe the media delivers false information with or without intention, thus potentially causing a vicious cycle of mistrust and alternative media usage.

The media are expected to provide the public with reliable information. Therefore, perceptions that the media is unintentionally or intentionally dishonest are likely to erode media trust (Pinkleton et al., 2012). Indeed, there is evidence which shows that increased concerns about the accuracy and reliability of political information are related to lower media trust (Van Aelst et al., 2017). It is important to note here that while PMI/PDI and media trust may seem related, they are theoretically distinct. For instance, PMI and PDI focus specifically on perceptions regarding the accuracy or intentional deception of news content. In contrast, media trust captures broader confidence in news organizations, journalists, and the institutional independence of the news media (see Newman et al., 2021). While trust focuses on dimensions such as credibility and reliability, PMI/PDI specifically focuses on perceptions, especially the attribution of intent in the case of PDI, which media trust measures do not directly assess. Although there are some correlational investigations, existing empirical evidence remains insufficient to definitively establish a causal relationship between media exposure and perceptions of news sources (Hameleers, Brosius, & de Vreese, 2022). Furthermore, since PMI/PDI call into question the reliability and accuracy of the media, individuals may turn to alternative media sources for information consumption (Prochazka & Schweiger, 2019; Tsfati & Cappella, 2003).

Consequences of Perceptions of Mis/Disinformation: The Role of Selective Exposure on News Media Use

When individuals perceive the news to be intentionally or unintentionally misleading, they may respond by limiting their use of such outlets. For instance, a recent study by Hameleers, Brosius, and de Vreese (2022) found that PMI/PDI were associated with lower traditional media news consumption and increased consumption of news on social media and non-mainstream alternative outlets. However, recent findings from public polls in the United States suggest that individuals are also concerned about mis- and disinformation in the news they encounter on social media (Boulianne & Hoffmann, 2024; Vigderman, 2024), particularly when it is aimed at disrupting democratic processes (Sanchez et al., 2022). While correlational, these findings suggest that there is a clear link between PMI/PDI and reduced traditional but also social media news use in the U.S. context.

Theoretically, the impact of these perceptions on media consumption can be understood through theories of selective exposure and motivated reasoning (Hart et al., 2009; Kunda, 1990). Foundational work in selective exposure research argues that individuals tend to select information that aligns with their pre-existing views over information that does not align with their views (Frey, 1986). For example, Garrett and Stroud (2014) found that regardless of political affiliation, individuals were more likely to select congenial news stories, although only Republicans were more likely to also avoid uncongenial news stories compared to individuals with other political affiliations. This congenial bias is likely due to one’s motivations to defend one’s own attitudes and hold accurate beliefs (Hart et al., 2009). Indeed, studies have found that concerns about misinformation on traditional and social media lead to selective exposure (Harris et al., 2024; Wu-Ouyang, 2024). Individuals may perceive information that they consume as not biased, while the general “media” are biased (see Barnidge et al., 2020). Research from misinformation literature also suggests that individuals tend to verify headlines congruent with their political ideology (Chan et al., 2024; Edgerly et al., 2019).

Although both traditional and social media are seen as potential sources of misinformation and disinformation, they are distinct in terms of their structure, users’ consumption patterns, and effects. Traditional media typically operates under editorial scrutiny, and journalistic standards which are designed to limit the spread of false information. Because fake news websites reach only small segments of the public, most people may indirectly be confronted with such stories via mainstream news coverage, and those who trust these outlets are especially likely to accept the information they provide (Tsfati et al., 2020). Therefore, traditional outlets may paradoxically amplify false information during efforts to debunk or report on misinformation (Tsfati et al., 2020). By contrast, on social media, news spreads through peer-to-peer diffusion, allowing news, especially the false stories to travel farther, faster, and deeper than fact-checked content (Vosoughi et al., 2018). Empirical evidence shows that social network platforms are the leading conduit for disinformation, particularly during the election periods (Benaissa Pedriza, 2021). These structural advantages such as their popularity and massive use, ease of generation and dissemination, and high velocity make social media a particularly powerful source of fake news dissemination (Aïmeur et al., 2023).

Tsfati and Cappella (2003) found that individuals who have more distrust in news media are less likely to consume traditional media sources, such as television news and newspapers, and more likely to consume other media sources, such as talk radio or the internet. More recently, Fletcher and Park (2017) found that across 11 countries, greater distrust in news media was associated with increased use of non-traditional outlets as a main source of news, including online-only or digitally born sources. These findings indicate that individuals who do not trust mainstream news media may be more likely to seek out alternative sources. However, individuals increasingly also do not trust social media for news either (Karlsen & Aalberg, 2023). A study by Park et al. (2020) found that across 26 countries, social media news use was associated with lower trust in the news in general. Since the majority of users are concerned about being exposed to misinformation online (Knuutila et al., 2022), these individuals may focus on selecting sources they perceive are trustworthy. In line with past research on selective exposure and motivated reasoning (Hart et al., 2009; Kunda, 1990), users who have higher PMI/PDI may not seek out traditional and social media news outlets for information.

AI as Alternative News Media

Amid declining trust in traditional outlets, AI has become an increasingly common means of information seeking, with AI-powered tools now capable of aggregating and personalizing news content (Simon & Isaza-Ibarra, 2023). Advances in natural language processing and machine learning enable AI-powered platforms to streamline data collection, analysis, and news writing, offering fast and scalable news production. Advocates of automated journalism highlight AI’s potential for offering what some may view as more accurate and objective coverage of events (Graefe, 2016).

Previous research suggests that AI as a news source can reduce perceived human bias. For example, a study by Cloudy et al. (2022) found that attributing a news story to an AI journalist, rather than a human journalist, mitigated hostile media bias. Similarly, Waddell (2019) demonstrated that news attributed to an automated author was perceived as less biased and more credible compared to news attributed to a human author. These findings align with the concept of the machine heuristic, which posits that machine sources are viewed as more objective and systematic than humans (Sundar, 2008). The perceived objectivity of AI may appeal to those concerned about bias in the news. Therefore, we can expect that those with PMI/PDI in traditional and social media news outlets may turn to AI as an alternative.

A large body of research suggests that when people perceive low objectivity and trustworthiness of traditional news outlets, they often seek alternative sources of information. For instance, Fletcher and Park (2017) found that individuals with low trust in the traditional news media tend to favor non-mainstream sources, such as social media, blogs, and digital-born news providers to fulfill their information needs. In a qualitative study, Edgerly (2017) highlighted that a skeptical orientation toward news has led individuals to seek information from multiple sources, even after reading information from legacy outlets like the New York Times and the Washington Post, to verify and ensure that they have good, unbiased information about a topic or event.

Social media platforms, although they were once thought of as democratizing forces for news, have also faced significant barriers to maintaining accurate and objective information sources due to a large volume of misinformation shared through online social networks (Newman et al., 2022). Recent public polls have revealed that the percentage of social media news consumers in the U.S. who cite inaccuracy as their primary concern has risen from 31% to 40% over the past 5 years (Wang & Forman-Katz, 2024).

Together, these studies underscore that as distrust in traditional and social media news intensifies, especially amid growing concerns about mis- and disinformation, individuals may increasingly turn to AI for information seeking. AI-driven news aligns with these preferences, as AI has been perceived to be less biased (e.g., Waddell, 2019). Moreover, research from the Reuters Institute has found that 36% of individuals in the U.S. trust news from AI (Arguedas, 2024). As AI-based information and news enable personalization, meaning that the audience gets to seek out information they desire proactively, AI may further appeal to individuals who feel unfulfilled by existing news channels. Based on this overview, we propose the following hypotheses and research questions:

H1–2: Exposure to information about (a) misinformation and (b) disinformation in traditional media news will (H1) reduce intentions to consume traditional media news and (H2) increase intentions to consume AI news, compared to the accurate information or control conditions.

RQ1–2: What is the effect of exposure to information about (a) misinformation and (b) disinformation in social media news on intentions to consume (RQ1) social media news and (RQ2) AI news, compared to the accurate information or control conditions?

However, similar to the influence of PMI/PDI on traditional and social media news use, the reliance on AI as an alternative source of news could differ along ideological lines, especially in a politically charged context such as the U.S..

The Moderating Role of Political Ideology

In the U.S., public trust in the news media has declined. Recent longitudinal analyses of media trust have revealed that both Democrats and Republicans have reported lower trust in the media since 1972 (Brenan, 2024). This mistrust is correlated with a concern that misinformation and disinformation are present in the news across the ideological spectrum (i.e., liberals and conservatives; BBC, 2024). Furthermore, concerns about the presence of misinformation and disinformation extend to news on traditional and social media outlets (Sanchez et al., 2022; Vigderman, 2024). These concerns may result in lower media news use on traditional and social media for individuals, regardless of ideology. Therefore, it can be argued that the relationship between PMI/PDI and traditional/social media news consumption may be negative for both liberals and conservatives. However, a growing body of research suggests that the strength of this effect could vary between liberals and conservatives.

The relationship between news consumption and media trust has increasingly become split along ideological lines in the United States. For instance, recent public opinion polls suggest that only 40% of Republicans trust national news organizations, while 78% of Democrats reported trust in national news organizations (Eddy, 2024). There is also growing evidence that compared to liberals and Democrats, Republicans and conservatives are alienated in that they only trust a small minority of news outlets, such as Fox News (Jurkowitz et al., 2020). Existing research has argued that this could be due to the hostile media effect, which suggests that individuals perceive news media to be biased against their point of view even when the coverage is neutral (Kelly, 2019; Vallone et al., 1985). More specifically, the tendency to perceive news media as biased is higher among Republicans and conservatives (Hansen & Kim, 2011). Another explanation for this effect could be due to a perception among conservatives that the media is biased toward liberals (see Lee, 2005). This perceived bias has been amplified since the ascent of Donald Trump, who has repeatedly attacked non-conservative media and referred to them as the enemy and “fake news” (Meeks, 2020). In turn, conservative and Republican voters have been increasingly distrustful of mainstream media (see Mourão et al., 2018) and engage in selective exposure due to hostile media perceptions (Barnidge et al., 2020; Stroud, 2010). PMI/PDI likely exacerbate these perceptions and news consumption patterns. This is not to say that liberals or Democrats do not engage in similar behavior. However, based on existing research, we can expect the strength of these effects to be stronger for conservatives (Garrett & Bond, 2021; Nikolov et al., 2021; Rao et al., 2022).

Similarly, conservatives may be less likely to accept AI as a source for news. For example, Yang et al. (2023) found that conservative ideology was positively associated with AI risk perceptions and negatively associated with support for the use of AI. Furthermore, political conservatism has also been linked with aversion toward AI (Castelo & Ward, 2021). One explanation may be that conservative individuals may perceive AI as a liberal-biased media source. Recent work has found that responses from ChatGPT may indeed indicate a left-leaning bias (Hartmann et al., 2023). Conservatives may aim to approach news sources that are perceived to be more aligned with their existing views, rather than selecting a perceived left-leaning news source. Finally, since conservatives place less trust in scientists to develop AI compared to liberals (i.e., Yang et al., 2023), they may be more hesitant to use AI as an information source due to an overall lack of trust in scientists (Gauchat, 2012). Thus, we argue that conservatives are less likely to rely on AI as a news source compared to liberals. Specifically, we propose the following:

H3: Political affiliation will moderate the effect of exposure to information about (a) misinformation and (b) disinformation in traditional media news on intentions to consume traditional media news, with greater decreases for liberals and conservatives compared to individuals in the accurate information or control conditions.

RQ3–4: How will political affiliation moderate the effect of exposure to information about (a) misinformation and (b) disinformation in social media news on intentions to consume (RQ3) social media news and (RQ4) AI news?

H4: Political affiliation will moderate the effect of exposure to information about (a) misinformation and (b) disinformation in traditional media news on intentions to consume AI news, with larger increases for liberals than conservatives compared to the accurate information or control conditions.

Moderating Effects of Media Trust

In addition to political ideology, general trust in news media may interact with PMI and PDI to shape media news use. For instance, the impact of PMI and PDI on media news use may be driven by distrust in the news media. Existing research indicates that lower trust in news media is associated with reduced news consumption (Hameleers, Brosius, & de Vreese, 2022; Tsfati & Peri, 2006). Furthermore, existing research has found that distrust in mainstream news drives consumption of alternative news channels (Benkler et al., 2018). As such, we propose the following:

H5: Trust in traditional media will moderate the effect of exposure to information about (a) misinformation and (b) disinformation in traditional media news on intentions to consume AI news, with lower-trust individuals reporting higher intentions than higher-trust individuals compared to the accurate information and control conditions.

RQ5: How will trust in traditional media moderate the effect of exposure to information about (a) misinformation and (b) disinformation in social media news and intentions to consume AI news?

Here, we conducted a pre-registered experiment (https://osf.io/j59fe)¹ to examine whether PMI/PDI have a causal effect on traditional and social media news consumption, as well as AI news consumption. In addition, we investigated whether political ideology and media trust moderate these relationships (see Figure 1 for a visual representation).

Figure 1.

Conceptual Model and Summary of Hypotheses and Research Questions.

Method

Sample and Procedures

We conducted the pre-registered survey experiment in December 2024. A priori power analysis for F-tests (analysis of covariance [ANCOVA]) using G*POWER 3.1 (Faul et al., 2009) indicated that 612 participants would be required to detect a medium effect (f = 0.15) at an alpha of 0.80. We recruited 838 U.S. adults using CloudResearch’s Connect, which uses verification strategies to confirm that respondents are actual people (Hartmann et al., 2023). The sample was matched to the U.S. Census in terms of age, gender, and race/ethnicity. As online survey participant pools are often skewed toward Democrats/Liberals, we set political party quotas to reflect a roughly even number of self-reported Democrats and Republicans, in addition to Independents. We removed participants who failed two attention checks asking them to recall which type of media was discussed in the story they read and the name of the town (n = 201). The final sample consisted of 637 participants, out of which there were 294 (46.2%) self-reported Democrats, 249 Republicans (39.1%), and 94 Independents (14.8%). In total, 51% participants were males (n = 325). The participants’ ages ranged from 18 to 82 years (M_age= 45.98, SD_age = 15.80). In terms of race, most of the participants were White/Caucasian (67.2%), followed by Black/African American (13.3%), Hispanic/Latino (8.3%), Asian (6.8%), Multiracial (2.2%), Biracial (1.4%), and Others (0.6%). Regarding education, the majority of the participants held a bachelor’s degree (45.9%), followed by some college but no degree (17.5%), master’s degree (12.1%), high school diploma or lower (10.7%), associate’s degree (9.4%), and doctoral or professional degree (4.4%). Participants provided informed consent and were compensated $1.50 for participating. The study was approved by the institutional review board (IRB) from one of the authors’ institutions.

This study used a 3 (PDI vs. PMI vs. Accuracy) × 2 (traditional vs. social news media) between-subjects design, along with a control condition. Participants completed a pretest questionnaire, which included questions about political affiliation, ideology, and interest, as well as media trust and media use questions. Then, participants were randomly assigned to one of the seven conditions, including the control group. In each condition, participants read a story where the media (either traditional or social media) provided misinformation unintentionally, provided disinformation on purpose, or provided accurate information about a health crisis in the fictional city of Greenfield. In the control group, participants read a story about residents of Greenfield celebrating an annual harvest festival. Across all conditions, participants had to spend at least 30 seconds reading the story before they could proceed. Existing research has argued that factors such as the length of a message or how much evidence is provided can impact findings (Han & Fink, 2012). As such, all the stories in each condition and the control group were created to be identical in length (roughly between 160 and 182 words) and to have the same number of arguments (see Supplementary Materials, Appendix A, for the stimuli).

Manipulation Check

For the manipulation check, we asked participants to indicate what they thought the story was about (see Supplementary Materials, Appendix B, for response wordings). Results from a chi-square test of independence revealed that the manipulation was effective, χ²(18, N = 636) = 1390.17, p < .001, Cramér’s V = .85 (p < .001). A vast majority of those in the PMI × misinformation in traditional media (76.1%) and social media conditions (71.4%) correctly indicated that the story focused on the news media providing false information unintentionally. Likewise, a majority of participants in the PDI × disinformation in traditional media (94.7%) and social media conditions (90.7%) correctly indicated that the story focused on the news media providing false news intentionally and deliberately. Then, those in the accuracy × traditional (92.5%) and social media conditions (84%) correctly identified that the story was focused on the news media providing accurate information. Finally, 96.5% of those in the control group correctly identified that the story was focused on celebrating an annual harvest.

Measures

Pretest Measures

Political Ideology

Participants were asked to report their political ideology on a scale from 1 = Very liberal to 7 = Very conservative (see Table 1 for detailed descriptive statistics and inter-item correlations).

Table 1.

Zero-Order Correlation Between Key Variables With Descriptive Statistics.

Variables	M	SD	1	2	3	4	5	6	7
1. Traditional Media News^a	3.31	1.82	—
2. Social Media News^a	2.93	1.66	.197**	—
3. AI News^a	2.78	2.06	194**	472**	—
4. Traditional Media News^b	3.30	1.85	.930**	190**	194**	—
5. Social Media News^b	2.68	1.62	.189**	914**	456**	.188**	—
6. AI News^b	2.75	2.05	.224**	.510**	830**	.24**	.542**	—
7. Media Trust	3.63	1.55	.330**	167**	152**	336**	.171**	.154**	—

Note. Superscripts indicate the measurement times: ^arepresents pretest and ^brepresents post-test.

p < .01 (two-tailed).

Media Trust

We measured media trust by asking participants to indicate their agreement with five items on a 7-point Likert-type scale (1 = Strongly disagree to 7 = Strongly agree). Some items included, “I think you can trust the news most of the time,” and “The news media are independent from undue political or government influence most of the time” (see Supplementary Materials, Appendix B, for item wordings). Participant responses were averaged to create the media trust variable (Newman et al., 2021; M = 3.63, SD = 1.55, α = 0.91).

Post-test Measures

Media News Use Variables

Participants were asked to indicate how often they used television news, newspapers, Facebook, Twitter/X, Instagram, and TikTok to get news about social, political, or public matters in a typical week on a 7-point Likert-type scale ranging from 1 = Never to 7 = Six or seven days (see Ahmed & Rasul, 2022; Ahmed et al., 2022 for a similar approach). Responses for television news and newspapers were averaged to create the traditional media news use variable (M = 3.30, SD = 1.85, r = 0.37, p < .001, Spearman-Brown = 0.54). Responses for Facebook, Twitter/X, Instagram, and TikTok were averaged to create the social media news use variable (M = 2.68, SD = 1.62, α = 0.71). We asked participants to indicate their intentions to use the media for news in the next week.

AI News Use

Participants were asked to indicate how often they intend to use AI technologies to get news about social, political, or public matters in a typical week on a single-item 7-point Likert-type scale ranging from 1 = Never to 7 = Six or seven days (M = 2.75, SD = 2.05). Since people are generally not aware of AI technologies (Kennedy et al., 2023), we provided participants with examples of such technologies (i.e., ChatGPT, Microsoft Bard, Alexa, Siri, etc.).

Analytical Approach

We controlled for participants’ age, gender, political interest,² prior media news use, and AI news use in our analyses. To test the direct effects, we relied on ANCOVAs and corrected for multiple comparisons using Fisher’s Least Significant Difference (LSD).³ For the moderation analyses, we relied on general linear models (GLMs) with robust standard errors and maximum likelihood estimation. Furthermore, analyses of hypotheses and research questions related to traditional media news were conducted only on participants who were assigned to traditional media news conditions and the control condition (n = 359). Similarly, analyses about hypotheses and research questions related to social media news involved participants who were assigned to social media news conditions and the control conditions (n = 359).

Results

We conducted randomization checks to assess whether assignment to conditions was random. No significant differences were found between conditions for age, gender, or education, suggesting that randomization was significant.⁴

Direct Effects of Misinformation and Disinformation on News Consumption Intentions

First, we tested H1a and H1b. We found a significant main effect of traditional media conditions on future intentions to consume traditional media news, F(3, 355) = 4.328, p = .005, η_partial² = .036. The pairwise comparisons revealed that those in the PMI about traditional media news condition reported significantly lower intentions to consume traditional media news than those in the accuracy condition (p = .001), but not the control condition (p = .061), partially supporting H1a. Similarly, those in the PDI condition reported significantly lower intentions to consume traditional media than those in the accuracy condition (p = .006) but not the control condition (p = .175). There were no significant differences between both PMI/PDI and control conditions (see Figure 2). Therefore, H1b was partially supported.

Figure 2.

Traditional Media News Use Intentions by Experimental Conditions (H1).

For RQ1, there was a significant main effect of social media news conditions on intentions to consume social media news, F(3, 355) = 3.282, p = .021, η_partial² = .027. Pairwise comparisons revealed that those in the PMI about social media news condition reported lower intentions to consume social media news compared to the accuracy (p = .010) and control conditions (p = .046). Likewise, those in the PDI about social media news conditions reported lower intentions to consume social media news compared to the accuracy (p = .019) condition, but not the control condition (p = .077). Overall, PMI about social media news led to weaker intentions to consume social media news in the future compared to the accuracy and control conditions (see Figure 3). However, PDI about social media news led to weaker intentions to consume social media news only compared to the accuracy condition (RQ1). As such, RQ1 was partially significant.

Figure 3

AI News Use Intentions by Experimental Conditions (H2).

Next, we assessed H2a and H2b. There was a significant main effect of traditional media news conditions on intentions to consume AI news, F(3, 355) = 3.074, p = .028, η_partial² = .026. The pairwise comparisons revealed that those in the PMI condition reported lower intentions to consume AI news compared to the control condition (p = .039), but not the accuracy condition (p = .478). Furthermore, those in the PDI condition also reported lower intentions to consume AI news compared to the control condition (p = .018), but not the accuracy condition (p = .646). Specifically, PMI/PDI led to weaker intentions to consume AI news compared to the accuracy condition, in contrast to H2a and H2b (see Figure 4). Therefore, neither H2a nor H2b was supported (see Table 2 for means).

Figure 4

Social Media News Use Intentions by Experimental Conditions (RQ1).

Table 2

Means and Standard Deviations for Dependent Variables by Condition.

Condition	Future Intention to Use Traditional Media News	Future Intention to Use Social Media News	Future Intention to Use AI News
Control	3.18 (1.74)	2.75 (1.58)	3.11 (2.22)
PDI—Traditional Media News	3.14 (1.99)^a	—	2.75 (2.08)
PMI—Traditional Media News	3.05 (1.71)^a	—	2.65 (2.05)
Accuracy—Traditional Media News	3.50 (1.86)	—	2.57 (1.92)
PDI—Social Media News	—	2.65 (1.52)^{a b}	2.97 (2.13)
PMI—Social Media News	—	2.59 (1.46)^{a b}	2.49 (1.95)
Accuracy—Social Media News	—	2.78 (1.62)	2.72 (1.99)

Note. Comparisons are based on pairwise results at p < .05.

Indicates that it is significantly different from the accuracy condition. ^bIndicates that it is significantly different from the control condition.

For RQ2, there was a significant main effect of social media news conditions on intentions to consume AI news, F(3, 355) = 4.993, p = .002, η_partial² = .041. However, Levene’s test indicated a violation of the homogeneity of variance assumption (p < .05). Therefore, we conducted a GLM with robust standard errors and maximum likelihood estimation to test RQ2. We found a significant main effect of social media news conditions on intentions to consume AI news, Wald χ²(3) = 11.181, p = .011. Pairwise comparisons showed that those in the PMI conditions reported lower intentions to consume AI news compared to the control condition (p = .001) but not for the accuracy condition (p = .077). There were no significant differences between the PDI and control or accuracy conditions (see Figure 5). In short, PMI about social media news leads to weaker intentions to consume AI news compared to the control condition while there are no significant effects of PDI (RQ2). Hence, RQ2 was partially significant.

Figure 5.

AI News Use Intentions by Experimental Conditions (RQ2).

Moderating Effects of Political Ideology

Regarding H3a and H3b, and RQ3, we did not find a significant interaction between traditional media news conditions and political ideology on intentions to consume traditional media news, Wald χ²(18) = 19.424, p = .366. Therefore, both H3a and H3b were not supported. Likewise, the interaction between social media news conditions and political ideology on intentions to consume social media news was not significant, Wald χ²(18) = 17.109, p = .516. Based on this, RQ3 was not significant.

We then examined H4a and H4b and RQ4. The interaction between traditional media news conditions and political ideology on intentions to consume AI news was marginally significant, Wald χ²(18) = 28.855, p = .050. However, none of the pairwise comparisons and conditional effects reached significance, rejecting H4a and H4b. We also did not find a significant interaction between social media news conditions and political ideology on intentions to consume AI news, Wald χ²(18) = 27.450, p = .071. Hence, RQ4 was not significant.

Moderating Effects of Media Trust

Finally, we tested H5a and H5b and RQ5. We did not find a significant interaction between traditional media news conditions and media trust on intentions to consume AI news, Wald χ²(3) = 2.650, p = .449. Therefore, both H5a and H5b were not supported. Likewise, the interaction between social media news conditions and media trust on intentions to consume AI news was not significant, Wald χ²(3) = 1.149, p = .765. As such, RQ5 was not significant (see Table 3 for a detailed summary about hypotheses and research questions).

Table 3.

Summary of Hypothesis and Research Question Testing Results.

Hypothesis/RQ	Relationship tested	Result
H1a/b	PMI/PDI in TM → Future intentions to use TM news	PMI: Partially supported; PDI: Partially supported
RQ1a/b	PMI/PDI in SM → Future intentions to use SM news	PMI: Significant; PDI: Partially significant
H2a/b	PMI/PDI in TM → Future intentions to use AI News	Not supported
RQ2a/b	PMI/PDI in SM → Future intentions to use AI news	PMI: Supported; PDI: Not supported
H3a/b	Political ideology × PMI/PDI (TM) → Future intentions to use TM news	Not supported
RQ3a/b	Political ideology × PMI/PDI (SM) → Future intentions to use SM news	Not significant
H4a/b	Political ideology × PMI/PDI (TM) → Future intentions to use AI news	Not supported
RQ4a/b	Political ideology × PMI/PDI (SM) → Future intentions to use AI news	Not significant
H5a/b	Trust in TM × PMI/PDI (TM) → Future intentions to use AI news	Not supported
RQ5a/b	Trust in TM × PMI/PDI (SM) → Future intentions to use AI news	Not significant

Note. PMI = perceived misinformation; PDI = perceived disinformation; SM = social media; TM = traditional media. Hypotheses/research questions associated with political ideology and media trust represent predictions about moderation effects.

Discussion

This study provides insights into how PMI/PDI plays a significant role in shaping news consumption intentions across various media platforms. As expected, exposure to perceptions of PMI/PDI in traditional media led to a decrease in future intentions to consume traditional media news. This effect was particularly notable when compared to the accuracy condition. In a similar vein, PMI/PDI about social media as a news outlet lowered intentions to consume news from these platforms in the future. This reduction was evident when compared to both accuracy and control conditions, suggesting the negative impact of untruthful information perceptions on social media news consumption.

PMI/PDI in the traditional media also affected the intention to consume AI-generated news but in the negative direction. Notably, these perceptions led to lower consumption intentions compared to the control condition but not compared to the accuracy condition. Finally, the experiment investigated the potential moderating role of political ideology and media trust on the relationship between PMI/PDI and media consumption intentions: We did not observe a significant interaction effect among the suggested variables despite our predictions.

These findings are contrary to theories of selective exposure and motivated reasoning (Hart et al., 2009; Kunda, 1990), which contend that individuals are likely to seek and interpret information that is congruent with their pre-existing beliefs. However, our findings revealed that PMI/PDI causally reduced news use intention regardless of the source. While prior research has found that as PMI/PDI increase, individuals may turn to alternative sources for news (Hameleers, Brosius, & de Vreese, 2022),we found that these perceptions decrease social media and AI news consumption intentions. This reflects contagious cynicism, which originally theorized that cynicism about politicians spilled over to the media, resulting in lower trust (Cappella & Jamieson, 1997). In this case, PMI/PDI in traditional and social media news could make audiences cynical, which may extend to alternative news sources such as social media and AI (see Hopmann et al., 2015). Rather than viewing AI news as an alternative, individuals may perceive all mediated news as potentially compromised, reflecting a general erosion of trust in the news media. Ultimately, this raises questions regarding which sources individuals with higher PMI/PDI may turn to for news.

Political ideology did not moderate the relationship between PMI/PDI and traditional, social media, or AI news use in our experiment. One possible explanation is that specific content or attributes of different news media lead individuals to perceive those sources as biased. For example, our experimental stimuli were strictly controlled between conditions and provided no cues regarding the news source’s partisanship. In a less controlled setting, it is possible that manipulating the news source would introduce a partisan cue, which can function as a cognitive shortcut (Fiske & Taylor, 1991) and lead to different effects among liberals and conservatives. Existing research has shown that partisan cues lead to selective exposure (Stroud, 2010), or news avoidance depending on whether the cue is in-party or out-party (Mukerjee & Yang, 2021). Future work could expand on the role of political ideology, as well as explore the relationship between PMI/PDI and partisan news content over time.

We also did not find a moderating effect of media trust on the relationship between PMI/PDI and news use intentions. One explanation for this could be that media trust develops or erodes over time (see Hanitzsch et al., 2018), and therefore, exposure to a news story at one time is likely not enough to activate differences based on trust (i.e., reinforcing spirals; Slater, 2007). This is further reflected in our findings, as, notably, prior news consumption (measured in the pretest) was identified as a stronger predictor of media behavior than our experimental manipulations. Furthermore, this suggests that existing low levels of news consumption could likely be exacerbated by PMI/PDI. This aligns with past research highlighting a broader trend of disengagement from news consumption. For example, Wojcieszak et al. (2024) found that approximately only 3% of individuals’ browsing histories were dedicated to news websites. These results are concerning, as democracy is dependent on informed citizens. This highlights the necessity for strategic interventions that not only address distrust of news media and PMI/PDI but also foster engagement with a broad spectrum of media outlets to reinforce the foundations of democratic governance.

One way to increase news trust and lower PMI/PDI could be to highlight the accuracy of news in the media. For instance, we found that when individuals in the accuracy conditions reported higher intentions to consume traditional and social media news. This falls in line with prior work that examines the role of accuracy motivations when interpreting and detecting misinformation (Albarracín, 2021; Chan, 2024; Rathje et al., 2023). Another way to increase public trust in the news media could be through transparency cues. Recent work has argued that increased transparency about an outlet’s partisan and ideological leanings may increase trust in a limited way, especially when an outlet does not lean toward one group or side (DuBosar et al., 2025). This further highlights a general distrust of traditional media news and should prompt future action to restore the credibility of traditional media news. In addition, future research should examine how PMI/PDI in AI news may impact future intentions to adopt AI news as an information source, consistent with accuracy-based motivations for selective exposure (Hart et al., 2009).

Our findings also have real-world implications for news organizations and policy makers, who are struggling to restore public trust in news. First, our findings demonstrate that increasing perceptions that the media is spreading mis- and disinformation make individuals hesitant to consume news from traditional media, social media, and AI. Researchers have long argued that the media’s role is to inform citizens in an ideal deliberative democracy (Strömbäck, 2005). However, PMI and PDI could significantly weaken the ability of the media to inform citizens. Furthermore, while skepticism of social media news has been well documented, our results indicate that PMI and PDI also suppress intentions to consume news through AI. This suggests that increasing efforts by newsrooms to integrate AI in their production may be met with resistance among the public. For example, recent work suggests that while AI disclosures by news outlets spark curiosity, they also result in lower trust and credibility (Toff & Simon, 2024). Therefore, interventions aimed at rebuilding media trust should not only address PMI/PDI but also take into account how newer technologies such as AI are perceived by individuals.

Limitations and Future Directions

One limitation of this study is the reliance on an experimental design to manipulate PMI and PDI in traditional and social media news. While this approach allowed for causal inferences, the observed effect size was small. This may be because trust in media channels develops over time rather than from a single exposure (Slater, 2007); an experimental stimulus may therefore only have a limited effect. Therefore, future studies may consider employing longitudinal designs using digital trace data to track changes in trust and capture news media use patterns. Another limitation of our study is that we rely on self-reported measures of media use, which are often underreported or exaggerated in some instances (Prior, 2009). In addition, while we distinguish between traditional and social media news use at a broad level, we do not account for the online presence of news outlets (e.g., following a TV or newspaper’s account on social media) or differentiate between curated versus incidental exposure to news on social media. Future research could benefit from measuring platform-specific news use more precisely. Combining survey data with novel techniques, such as tracking web browsing data, could also provide a deeper understanding of individuals’ actual media diets and help capture incidental exposure more accurately (Wojcieszak et al., 2024). Moreover, while our study operationalized and conceptualized AI news use, more work is needed in this area. Future research could benefit from a more in-depth conceptualization and operationalization of what AI news use encompasses. For instance, is it AI-curated or completely AI-generated news? Our experiment was also highly controlled to isolate the effects of PMI/PDI and to avoid any confounds. However, as with most tightly controlled experiments, this may have resulted in a tradeoff between internal and external validity.

Through an experiment of U.S. adults, we find that PMI/PDI generally reduces news media consumption. While our study focused on the potential for AI news as an alternative source that people may gravitate toward, our experiment demonstrated that PMI/PDI in traditional and social media reduces individuals’ intentions to use AI news as well.

Footnotes

Data Availability Statement

The data for this study will be made available upon reasonable request.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Funding

The author(s) received no financial support for the research,authorship,and/or publication of this article.

ORCID iDs

Muhammad Ehab Rasul

Christopher Calabrese

Yoo Jung Oh

Hee Jung Cho

Moonsun Jeon

Mark Boukes

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Muhammad Ehab Rasul (PhD,University of California,Davis) is a research fellow at the Centre for Trusted Internet and Community at the National University of Singapore. His research interests focus on the impact of digital and emerging media,misinformation,and information consumption on political attitudes and behavior.

Christopher Calabrese (PhD,University of California,Davis) is an Assistant Professor in the Department of Communication at Clemson University and a Faculty Scholar at the Clemson University School of Health Research. His research examines the effects and processes through which persuasive messaging and online technologies influence human behaviors and decision-making.

Yoo Jung Oh (PhD,University of California,Davis) is an Assistant Professor in the Department of Communication at Michigan State University. She specializes in artificial intelligence (AI) dialogue systems,commonly referred to as “AI chatbots.” With an interest in the dynamics of human–AI communication,her research investigates the impact of persuasive AI on human perceptions and behaviors.

Hee Jung Cho (MA,Ewha Woman’s University) is a doctoral student in the Department of Communication at Michigan State University. Her current research interest lies in examining the affective and cognitive mechanisms in information processing,particularly in the context of health and risk communication.

Moonsun Jeon (MA,University of Georgia) is a doctoral student in the Department of Communication. Her research interests lie in persuasion and social influence,specifically in the field of health and environmental communication.

Mark Boukes (PhD,University of Amsterdam) is an Associate Professor in the department of Communication Science at the University of Amsterdam. His research interests include,but are definitively not limited to,journalism,media effects,infotainment,and (digital) research methods.

References

Ahmed

Rasul

M. E.

(2022). Social media news use and COVID-19 misinformation engagement: Survey study. Journal of Medical Internet Research, 24(9), e38944. https://doi.org/10.2196/38944

Ahmed

Rasul

M. E.

Cho

(2022). Social media news use induces COVID-19 vaccine hesitancy through skepticism regarding its efficacy: A longitudinal study from the United States. Frontiers in Psychology, 13, 1–7. https://doi.org/10.3389/fpsyg.2022.900386

Aïmeur

Amri

Brassard

(2023). Fake news, disinformation and misinformation in social media: A review. Social Network Analysis and Mining, 13(30), 1–36. https://doi.org/10.1007/s13278-023-01028-5

Albarracín

(2021). Conspiracy beliefs: Knowledge, ego defense, and social integration in the processing of fake news. In Greifeneder

Jaffé

M. E.

Newman

E. J.

Schwarz

(Eds.), The psychology of fake news: Accepting, sharing, and correcting misinformation (pp. 196–219). Routledge/Taylor & Francis Group. https://doi.org/10.4324/9780429295379-14

Arguedas

A. R.

(2024, June 17). Public attitudes towards the use of AI in journalism. Reuters Institute for the Study of Journalism. https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2024/public-attitudes-towards-use-ai-and-journalism

Bachmann

Valenzuela

(2023). Studying the downstream effects of fact-checking on social media: Experiments on correction formats, belief accuracy, and media trust. Social Media + Society, 9(2), 1–13. https://doi.org/10.1177/2056305123117969

Barnidge

Gunther

A. C.

Kim

Hong

Perryman

Tay

S. K.

Knisely

(2020). Politically motivated selective exposure and perceived media bias. Communication Research, 47(1), 82–103. https://doi.org/10.1177/0093650217713066

Benaissa Pedriza

. (2021). Sources, channels and strategies of disinformation in the 2020 US election: Social networks, traditional media and political candidates. Journalism and Media, 2(4), 605–624. https://doi.org/10.3390/journalmedia2040036

Benkler

Farris

Roberts

(2018). Network propaganda: Manipulation, disinformation, and radicalizaton in American politics (Vol. 1). Oxford University Press. https://doi.org/10.1093/oso/9780190923624.001.0001

10.

Boulianne

Hoffmann

C. P.

(2024). Perceptions and concerns about misinformation on Facebook in Canada, France, the US, and the UK. International Journal of Public Opinion Research, 36(4), 1–12. https://doi.org/10.1093/ijpor/edae048

11.

Brenan

(2024, October 14). Americans’ trust in media remains at trend low. Gallup. https://news.gallup.com/poll/651977/americans-trust-media-remains-trend-low.aspx

12.

Cappella

J. N.

Jamieson

K. H.

(1997). Spiral of cynicism: The press and the public good. Oxford University Press.

13.

Castelo

Ward

A. F.

(2021). Conservatism predicts aversion to consequential Artificial Intelligence. PLOS ONE, 16(12), e0261467. https://doi.org/10.1371/journal.pone.0261467

14.

Chan

(2024). Verification behaviors and countermeasures in the age of misinformation. Journalism & Mass Communication Quarterly, 101(1), 13–19. https://doi.org/10.1177/10776990231223998

15.

Cloudy

J. T.

Banks

J. T.

Bowman

N. D.

(2022). AI journalists and reduction of perceived hostile media bias: Replication and extension considering news organization cues. Technology, Mind, and Behavior, 3(3), 1–15. https://doi.org/10.1037/tmb0000083

16.

DuBosar

Hmielowski

J. D.

Rasul

M. E.

(2025). In Diversity We Trust? Examining the Effect of Political Newsroom Diversity on Media Trust, Use, and Avoidance. Communication Research. Advance online publication. https://doi.org/10.1177/00936502251341081

17.

Eddy

(2024, October 16). Republicans, young adults now nearly as likely to trust info from social media as from national news outlets. Pew Research Center. https://www.pewresearch.org/short-reads/2024/10/16/republicans-young-adults-now-nearly-as-likely-to-trust-info-from-social-media-as-from-national-news-outlets/

18.

Edgerly

(2017). Seeking out and avoiding the news media: Young adults’ proposed strategies for obtaining current events information. Mass Communication and Society, 20(3), 358–377. https://doi.org/10.1080/15205436.2016.1262424

19.

Edgerly

Mourão

R. R.

Thorson

Tham

S. M.

(2019). When do audiences verify? How perceptions about message and source influence audience verification of news headlines. Journalism & Mass Communication Quarterly, 97(1), 52–71. https://doi.org/10.1177/1077699019864680

20.

Egelhofer

J. L.

Lecheler

(2019). Fake news as a two-dimensional phenomenon: A framework and research agenda. Annals of the International Communication Association, 43(2), 97–116. https://doi.org/10.1080/23808985.2019.1602782

21.

Faul

Erdfelder

Buchner

Lang

A. G.

(2009). Statistical power analyses using G* Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149–1160. https://doi.org/10.3758/brm.41.4.1149

22.

Fiske

S. T.

Taylor

S. E.

(1991). Social cognition (2nd ed.). McGraw-Hill.

23.

Fletcher

Park

(2017). The impact of trust in the news media on online news consumption and participation. Digital Journalism, 5(10), 1281–1299. https://doi.org/10.1080/21670811.2017.1279979

24.

Frey

(1986). Recent research on selective exposure to information. Advances in Experimental Social Psychology, 19, 41–80. https://doi.org/10.1016/S0065-2601(08)60212-9

25.

Garrett

R. K.

Bond

R. M.

(2021). Conservatives’ susceptibility to political misperceptions. Science Advances, 7(23), eabf1234. https://doi.org/10.1126/sciadv.abf1234

26.

Garrett

R. K.

Stroud

N. J.

(2014). Partisan paths to exposure diversity: Differences in pro-and counterattitudinal news consumption. Journal of Communication, 64(4), 680–701. https://doi.org/10.1111/jcom.12105

27.

Gauchat

(2012). Politicization of science in the public sphere: A study of public trust in the United States, 1974 to 2010. American Sociological Review, 77(2), 167–187. https://doi.org/10.1177/0003122412438225

28.

Graefe

(2016). Guide to automated journalism. Tow Center for Digital Journalism, Columbia University. https://doi.org/10.7916/D80G3XDJ

29.

Gunther

A. C.

McLaughlin

Gotlieb

M. R.

Wise

(2017). Who says what to whom: Content versus source in the hostile media effect. International Journal of Public Opinion Research, 29(3), 363–383. https://doi.org/10.1093/ijpor/edw009

30.

Hameleers

(2022). “I don’t believe anything they say anymore!” Explaining unanticipated media effects among distrusting citizens. Media and Communication, 10(3), 158–168. https://doi.org/10.17645/mac.v10i3.5307

31.

Hameleers

Brosius

de Vreese

C. H.

(2022a). Whom to trust? Media exposure patterns of citizens with perceptions of misinformation and disinformation related to the news media. European Journal of Communication, 37(3), 237–268. https://doi.org/10.1177/02673231211072667

32.

Hameleers

Brosius

Marquart

Goldberg

A. C.

Van Elsas

De Vreese

C. H.

(2022b). Mistake or manipulation? Conceptualizing perceived mis-and disinformation among news consumers in 10 European countries. Communication Research, 49(7), 919–941. https://doi.org/10.1177/0093650221997719

33.

Han

Fink

E. L.

(2012). How do statistical and narrative evidence affect persuasion? The role of evidentiary features. Argumentation and Advocacy, 49(1), 39–58. https://doi.org/10.1080/00028533.2012.11821779

34.

Hanitzsch

Van Dalen

Steindl

(2018). Caught in the nexus: A comparative and longitudinal analysis of public trust in the press. The International Journal of Press/Politics, 23(1), 3–23. https://doi.org/10.1177/1940161217740695

35.

Hansen

G. J.

Kim

(2011). Is the media biased against me? A meta-analysis of the hostile media effect research. Communication Research Reports, 28(2), 169–179. https://doi.org/10.1080/08824096.2011.565280

36.

Harris

E. A.

DeMora

S. L.

Albarracín

(2024). The consequences of misinformation concern on media consumption. Harvard Kennedy School Misinformation Review, 5(3), 1–20. https://doi.org/10.37016/mr-2020-149

37.

Hart

Albarracín

Eagly

A. H.

Brechan

Lindberg

M. J.

Merrill

(2009). Feeling validated versus being correct: A meta-analysis of selective exposure to information. Psychological Bulletin, 135(4), 555–588. https://doi.org/10.1037/a0015701

38.

Hartmann

Schwenzow

Witte

(2023). The political ideology of conversational AI: Converging evidence on ChatGPT’s pro-environmental, left-libertarian orientation. arXiv. https://doi.org/10.48550/arXiv.2301.01768

39.

Hopmann

D. N.

Shehata

Strömbäck

(2015). Contagious media effects: How media use and exposure to game-framed news influence media trust. Mass Communication & Society, 18(6), 776–798. https://doi.org/10.1080/15205436.2015.1022190

40.

Jurkowitz

Mitchell

Shearer

Walker

(2020, January 24). U.S. media polarization and the election: A nation divided 2020. Pew Research Center. https://www.pewresearch.org/journalism/2020/01/24/u-s-media-polarization-and-the-2020-election-a-nation-divided/

41.

Karlsen

Aalberg

(2023). Social media and trust in news: An experimental study of the effect of Facebook on news story credibility. Digital Journalism, 11(1), 144–160. https://doi.org/10.1080/21670811.2021.1945938

42.

Kelly

(2019). Evaluating the news:(Mis) perceptions of objectivity an d credibility. Political Behavior, 41(2), 445–471. https://doi.org/10.1007/s11109-018-9458-4

43.

Kennedy

Tyson

Saks

(2023, February 15). Public awareness of artificial intelligence in everyday activities. Pew Research Center. https://www.pewresearch.org/science/2023/02/15/public-awareness-of-artificial-intelligence-in-everyday-activities/

44.

Knuutila

Neudert

L.-M.

Howard

P. N.

(2022). Who is afraid of fake news? Modeling risk perceptions of misinformation in 142 countries. Harvard Kennedy School Misinformation Review, 3(3), 1–13. https://doi.org/10.37016/mr-2020-97

45.

Kunda

(1990). The case for motivated reasoning. Psychological. Bulletin, 108(3), 480–498. https://doi.org/10.1037/0033-2909.108.3.480

46.

Lee

T.-T.

(2005). The liberal media myth revisited: An examination of factors influencing perceptions of media bias. Journal of Broadcasting & Electronic Media, 49(1), 43–64. https://doi.org/10.1207/s15506878jobem4901_4

47.

Lindgren

Lindholm

Vliegenthart

Boomgaarden

H. G.

Damstra

Strömbäck

Tsfati

(2024). Trusting the facts: The role of framing, news media as a (trusted) source, and opinion resonance for perceived truth in statistical statements. Journalism & Mass Communication Quarterly, 101(4), 981–1004. https://doi.org/10.1177/10776990221117117

48.

The majority of Americans are concerned about misinformation in the news, according to new research. (2024, October 29). BBC. https://www.bbc.co.uk/mediacentre/articles/2024/misinformation-in-the-news-research/

49.

Meeks

(2020). Defining the enemy: How Donald Trump frames the news media. Journalism & Mass Communication Quarterly, 97(1), 211–234. https://doi.org/10.1177/1077699019857676

50.

Mourão

R. R.

Thorson

Chen

Tham

S. M.

(2018). Media repertoires and news trust during the early Trump administration. Journalism Studies, 19(13), 1945–1956. https://doi.org/10.1080/1461670x.2018.1500492

51.

Mukerjee

Yang

(2021). Choosing to avoid? A conjoint experimental study to understand selective exposure and avoidance on social media. Political Communication, 38(3), 222–240. https://doi.org/10.1080/10584609.2020.1763531

52.

Newman

E. J.

Swire-Thompson

Ecker

U. K. H.

(2022). Misinformation and the sins of memory: False-belief formation and limits on belief revision. Journal of Applied Research in Memory and Cognition, 11(4), 471–477. https://doi.org/10.1037/mac0000090

53.

Newman

Fletcher

Schulz

Andi

Robertson

C. T.

Nielsen

R. K.

(2021, July 8). Reuters Institute digital news report 2021. Reuters Institute for the Study of Journalism. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3873260#paper-references-widget

54.

Nikolov

Flammini

Menczer

(2021). Right and left, partisanship predicts (asymmetric) vulnerability to misinformation. Harvard Kennedy School Misinformation Review, 1(7), 1–13. https://doi.org/10.37016/mr-2020-55

55.

Park

Fisher

Flew

Dulleck

(2020). Global mistrust in news: The impact of social media on trust. The International Journal on Media Management, 22(2), 83–96. https://doi.org/10.1080/14241277.2020.1799794

56.

Pinkleton

B. E.

Austin

E. W.

Zhou

Willoughby

J. F.

Reiser

(2012). Perceptions of news media, external efficacy, and public affairs apathy in political decision making and disaffection. Journalism & Mass Communication Quarterly, 89(1), 23–39. https://doi.org/10.1177/1077699011428586

57.

Prior

(2009). The immensely inflated news audience: Assessing bias in self-reported news exposure. Public Opinion Quarterly, 73(1), 130–143. https://doi.org/10.1093/poq/nfp002

58.

Prochazka

Schweiger

(2019). How to measure generalized trust in news media? An adaptation and test of scales. Communication Methods and Measures, 13(1), 26–42. https://doi.org/10.1080/19312458.2018.1506021

59.

Rao

Morstatter

Lerman

(2022). Partisan asymmetries in exposure to misinformation. Scientific Reports, 12(1), 1–11. https://doi.org/10.1038/s41598-022-19837-7

60.

Rasul

M. E.

Ahmed

Cho

Gil-Lopez

(2025). From podcasts to protests: Examining the influence of podcasts and misinformation on contentious political participation. Journal of Broadcasting & Electronic Media, 69(3), 273–293. https://doi.org/10.1080/08838151.2025.2495803

61.

Rathje

Roozenbeek

Van Bavel

J. J.

Van Der Linden

(2023). Accuracy and social motivations shape judgements of (mis) information. Nature Human Behaviour, 7(6), 892–903. https://doi.org/10.1038/s41562-023-01540-w

62.

Russell

S. J.

Norvig

(2010). Artificial intelligence: A modern approach (3d ed.). Prentice Hall.

63.

Rutherford

(2011). Anova and Ancova: A GLM approach (2nd ed.). Wiley. https://doi.org/10.1002/9781118491683

64.

Rzepka

Berger

(2018, December). User interaction with AI-enabled systems: A systematic review of IS research [Paper presentation]. International Conference on Information Systems (ICIS) Conference 2018, San Francisco, California, US. https://aisel.aisnet.org/icis2018/general/Presentations/7

65.

Sanchez

G. R.

Middlemass

Rodriguez

(2022, July 26). Misinformation and eroding of public confidence in democracy. Policy Commons. https://coilink.org/20.500.12592/bxbvq9

66.

Simon

F. M.

Isaza-Ibarra

L. F.

(2023). AI in the news: Reshaping the information ecosystem? (p. 24). Oxford Internet Institute, University of Oxford. http://doi.org/10.5287/ora-dx865edma

67.

Slater

M. D.

(2007). Reinforcing spirals: The mutual influence of media selectivity and media effects and their impact on individual behavior and social identity. Communication Theory, 17(3), 281–303. https://doi.org/10.1111/j.1468-2885.2007.00296.x

68.

Strömbäck

(2005). In search of a standard: Four models of democracy and their normative implications for journalism. Journalism Studies, 6(3), 331–345. https://doi.org/10.1080/14616700500131950

69.

Stroud

N. J.

(2010). Polarization and partisan selective exposure. Journal of Communication, 60(3), 556–576. https://doi.org/10.1111/j.1460-2466.2010.01497.x

70.

Sundar

S. S.

(2008). The MAIN model: A heuristic approach to understanding technology effects on credibility. In Metzger

M. J.

Flanagin

A. J.

(Eds.), Digital media, youth, and credibility (pp. 73–100). The MIT Press.

71.

Toff

Simon

F. M.

(2024). “Or they could just not use it?”: The dilemma of AI disclosure for audience trust in news. The International Journal of Press/Politics. Advance online publication. https://doi.org/10.1177/19401612241308697

72.

Tsfati

Boomgaarden

H. G.

Strömbäck

Vliegenthart

Damstra

Lindgren

(2020). Causes and consequences of mainstream media dissemination of fake news: Literature review and synthesis. Annals of the International Communication Association, 44(2), 157–173. https://doi.org/10.1080/23808985.2020.1759443

73.

Tsfati

Cappella

J. N.

(2003). Do people watch what they do not trust? Exploring the association between news media skepticism and exposure. Communication Research, 30(5), 504–529. https://doi.org/10.1177/0093650203253371

74.

Tsfati

Peri

(2006). Mainstream media skepticism and exposure to sectorial and extranational news media: The case of Israel. Mass Communication & Society, 9(2), 165–187. https://doi.org/10.1207/s15327825mcs0902_3

75.

Vallone

R. P.

Ross

Lepper

M. R.

(1985). The hostile media phenomenon: Biased perception and perceptions of media bias in coverage of the Beirut massacre. Journal of Personality and Social Psychology, 49(3), 577. https://doi.org/10.1037//0022-3514.49.3.577

76.

Van Aelst

Strömbäck

Aalberg

Esser

De Vreese

Matthes

Stanyer

. (2017). Political communication in a high-choice media environment: A challenge for democracy? Annals of the International Communication Association, 41(1), 3–27. https://doi.org/10.1080/23808985.2017.1288551

77.

Vigderman

(2024, September 26). 90% of people claim they fact-check news stories as trust in media plummets. Security.org. https://www.security.org/digital-security/misinformation-disinformation-survey/

78.

Vosoughi

Roy

Aral

(2018). The spread of true and false news online. Science, 359(6380), 1146–1151. https://doi.org/10.1126/science.aap9559

79.

Vraga

E. K.

Tully

(2021). News literacy, social media behaviors, and skepticism toward information on social media. Information, Communication & Society, 24(2), 150–166. https://doi.org/10.1080/1369118X.2019.1637445

80.

Waddell

T. F.

(2019). Can an algorithm reduce the perceived bias of news? Testing the effect of machine attribution on news readers’ evaluations of bias, anthropomorphism, and credibility. Journalism & Mass Communication Quarterly, 96(1), 82–100. https://doi.org/10.1177/1077699018815891

81.

Wang

Forman-Katz

(2024, February, 7). Many Americans find value in getting news on social media, but concerns about inaccuracy have risen. Pew Research Center. https://www.pewresearch.org/short-reads/2024/02/07/many-americans-find-value-in-getting-news-on-social-media-but-concerns-about-inaccuracy-have-risen/

82.

Wojcieszak

Menchen-Trevino

Clemm von Hohenberg

de Leeuw

Gonçalves

Davidson

Gonçalves

(2024). Non-news websites expose people to more political content than news websites: Evidence from browsing data in three countries. Political Communication, 41(1), 129–151. https://doi.org/10.1080/10584609.2023.2238641

83.

Wu-Ouyang

(2024). Comparing media systems in western democracies: Examining the role of multiplatform news use and fake news concern on news engagement and selective exposure. Journalism & Mass Communication Quarterly, 101(1), 45–70. https://doi.org/10.1177/10776990231217482

84.

Yang

Krause

N. M.

Bao

Calice

M. N.

Newman

T. P.

Scheufele

D. A.

Xenos

M. A.

Brossard

(2023). In AI we trust: The interplay of media use, political ideology, and trust in shaping emerging AI attitudes. Journalism & Mass Communication Quarterly, 102(2), 382–406. https://doi.org/10.1177/10776990231190868

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

0.00 MB

“It’s All Fake News!”: How Perceptions of Misinformation and Disinformation Influence News Consumption Across Traditional Media,Social Media,and AI

Abstract

Keywords

Literature Review

Understanding Perceptions of Mis/Disinformation

Consequences of Perceptions of Mis/Disinformation: The Role of Selective Exposure on News Media Use

AI as Alternative News Media

The Moderating Role of Political Ideology

Moderating Effects of Media Trust

Method

Sample and Procedures

Manipulation Check

Measures

Pretest Measures

Political Ideology

Media Trust

Post-test Measures

Media News Use Variables

AI News Use

Analytical Approach

Results

Direct Effects of Misinformation and Disinformation on News Consumption Intentions

Moderating Effects of Political Ideology

Moderating Effects of Media Trust

Discussion

Limitations and Future Directions

Footnotes

Data Availability Statement

Declaration of Conflicting Interests

Funding

ORCID iDs

Supplemental Material

Author Biographies

References

Supplementary Material