Drawing on Third-Person Effect (TPE) theory, this study investigates how exposure to warning labels on AI-generated disinformation shapes perceptual effects and behavioral consequences during the 2024 U.S. Presidential Election.

Drawing on Third-Person Effect (TPE) theory, this study investigates how exposure to warning labels on AI-generated disinformation shapes perceptual effects and behavioral consequences during the 2024 U.S. Presidential Election.
This study develops and tests the AI-Based Risk Information-Seeking Model (ARISM), which incorporates AI-related cognitive and affective factors, including privacy concerns and AI anxiety.
This research investigates the adoption, integration, and everyday use of social media by married, stay-at-home, conservative Muslim migrant women who followed their husbands from Turkey to the UK, living in Sussex.
Why do certain topics attract more public attention than others on digital platforms? The authors propose a dynamic cross-platform spillover explanation, examining how cross-platform trending influences public attention to political topics.
Drawing on Critical Technocultural Discourse Analysis of 1,500 TikTok videos, this study examines how non-Latinas and Latinas interact with ‘Latina makeup’ filters.
This paper offers a theoretical examination of AI-generated outputs, employing fact-checking as an epistemic lens.
This study employs the Fill-Mask Association Test in four principal Chinese Pre-trained language models (PLMs) with reference to the English PLMs.
This study examines the dynamics of cross-ideological communication across 11 major subreddits over 16 years. Going beyond purely textual elements, the authors analyse a broad range of possibilities for overlaps between subreddits.
This study examines the impact of cyberchondria on depressive symptoms using data from a nationally representative sample (N = 1,500) of Chinese adults.
Anand Badola reviews 'Safe havens for hate: the challenge of moderating online extremism’ by Tamar Mitts. The core argument focuses on the difference in moderation policies across different platforms that creates safe havens for extremist actors.
Drawing on interviews with members of Brazil’s Congress, this study investigates how politicians incorporate audience analytics into their communication strategies.
Applying the Draw-a-Scientist Test framework, the authors analyzed stereotypical representations of scientists in images generated by DALL·E 3 and examined potential amplification by benchmarking against Google Images.
Caravan - Ultraviolet Violence is Out Now! stream #instrumental #music #newrelease #outnow #Occult #newmusic #musicvideo #HardRock #MusicStreaming #art #NewMusic2026 #peaceful #instrumentals #recordlabel #caravan #newsong #NewMusicAlert #halloween #halloween2026 #happyhalloween #happyhalloween2026
🔥 FORCED DIRECT IS OUT NOW.
Take back control. No more waiting.
🎵 TheFallenAngel
#ForcedDirect #TheFallenAngel #OutNow #NewMusicFriday #RapRock
Sound Of Underground remix out now! 😎
M.A.N - Sound Of Underground (AxelPolo & Matt Rodgers Remix) [Allsense Recordings] 🔥
Now available to buy/stream on all platforms 🙌
Esa girl is so dangerous
soundcloud.com/notpirate/es...
#SoundCloud #music #notpirate #original
#newmusic #outnow #underground
#twitch #kick #hiphop #dangerous
Se viene mi primer Álbum Cyberpunk
in progress...
un adelanto
soundcloud.com/notpirate/ze...
#SoundCloud #music #notpirate #original
#newmusic #outnow #underground
#twitch #kick #cyberpunk
And So the Tension Raises...
Jet Trash is streaming now on Tubi
#tubi #jettrash #movie #crime #thriller #fight #osyikhile #sofiaboutella #robertsheehan #outnow #trippy #gangs #crimethriller #romance #cops #crookedcops #danger #mob #mysticism #tubimovies #jasperpaakkonen