#infodemic, one year on. News orgs are the most widely used source of information about coronavirus and have become even more central because-while overall reach has declined compared to earlier in the pandemic-reach of other sources has declined more. reutersinstitute.politics.ox.ac.uk/ongoing-infode… 1/9
Trust in news orgs declined by an average of 8 percentage points and trust in national governments has declined by average of 13pp. In most countries covered, national health authorities, global health authorities, and scientists+doctors+experts remain highly+broadly trusted 2/9
The 'trust gap' between coronavirus information from news organisations and information on different kinds of platforms remains pronounced. On average, gap btw news orgs and social media is 21pp, btw news and video sites 22pp, and btw news and messaging applications 28pp. 3/9
In terms of sources of false or misleading information about COVID-19, public concern is primarily centred on political actors. On average, 35% of respondents across eight countries say they have seen a lot or a great deal of false or misleading information from politicians 4/9
In terms of platforms, public concern over false or misleading information about COVID-19 is centred on social media. On average, 30% across eight countries say they think they have seen ‘a lot’ or ‘a great deal’ of false or misleading info about coronavirus on social media 5/9
Encouragingly, belief in misinformation about coronavirus vaccines is very low. In most countries, more than 90% of respondents surveyed do not believe ANY of the five false claims we include in the survey. But significant numbers answer "don't know", documenting uncertainty 6/9
Controlling for other factors, we find using news organisations as a source for news and information about coronavirus is significantly associated with lower belief in vaccine misinformation in all eight countries covered. (We found similar positive association last year) 7/9
In contrast, relying on messaging apps, social media, video sites for info about coronavirus is significantly associated with higher misinformation belief in most of the countries covered. (Last year we found no consistent, significant associations w/coronavirus knowledge.) 8/9
Report based on online survey of nationally representative sample in 🇦🇷🇧🇷🇩🇪🇯🇵🇪🇸🇰🇷🇬🇧🇺🇸
Even as the % who say they get news via Facebook continues to decline, a range of other social, video, and messaging platforms are growing in importance for discovery, many focused on on-site video, visuals, and more private experiences. Challenging environment for publishers 2/9
Generally, many of our respondents say they find it at least somewhat easy to tell trustworthy and untrustworthy news and information apart on various platforms, but there are real differences, with more concerned about how to navigate information on e.g. TikTok, X, Facebook 3/9
What might an AI-mediated information ecosystem look like?
Shuwei Fang @OpenSociety & @StructStories asked for scenarios. Mine, on interplay btw AI pragmatism, AI experimentalism, & AI incrementalism, draws on @risj_oxford research & more
My starting point? Public uptake will be one of the most important driving forces shaping the AI-mediated information ecosystem and, by extension, journalism & news media’s place in it. Demand is sometimes overlooked in discussions that tend to focus on actors on supply side 2/14
Extrapolating from public approach to previous digital technologies a likely approach might be "AI pragmatism", combination of (a) abstract concern about impact, (b) scepticism towards many of the institutions using AI, & (c) a practical appreciation of many AI applications 3/14
AI cannot reliably identify false news (let alone lies), despite what sales reps and boosters may claim. Too many false positives, false negatives, issues of bias, (let alone perceiving "entire meanings")
Because so much of the most potentially consequential misinfo, including false news and lies, is fundamentally political, there is real and perceived conflict of interests when govs' want to play role as arbiters of truth.
This is even more pronounced in low-trust contexts. 2/4
For years, experts have argued govs' and public authorities' most constructive role is indirect - convening whole-of-society responses and providing funding for independent fact-checkers, journalists, researchers, civil society
Second, what is crucial is not volume but influence. As @hugoreasoning and others have pointed out, attempts at mass persuasion mostly fail! . But one thing that often influence people is elite cues from politicians they support 3/7press.princeton.edu/books/hardcove… cambridge.org/core/books/nat…
Beyond inadverdently disseminating disinfo bcs of business-as-usual editorial practices, there are also parts of the media (e.g. some pundits, broadcast hosts) who are parts of what @sobieraj called "outrage industry" - even when working for news media global.oup.com/academic/produ… 2/6
Whether as sources (for news reporters), guests/subjects (for hosts and pundits), or important users and advertiseres (for platforms) - or just doing their own thing - some domestic political elites sometimes contribute to mis- and disinfo problems academic.oup.com/book/26406 3/6