Short panel presentation given in the context of the AI4EU WebCafe "The COVID-19 and Contact Tracing Apps" on June 23rd 2020, focusing on the problem of COVID-19 misinformation and how this could potentially affect the adoption of contact tracing apps.
Scanning the Internet for External Cloud Exposures via SSL Certs
COVID-19 Infodemic vs Contact Tracing
1. COVID-19 Infodemic vs
Contact Tracing
Dr. Symeon (Akis) Papadopoulos
Senior Researcher
Centre for Research and Technology Hellas (CERTH)
AI4EU Café on “COVID-19 and Contact Tracing Apps”
2. The COVID-19 Infodemic
• Unprecedented scale, variety and velocity of
information, including misinformation
• Complex (mis)information ecosystem
comprising politicians, established media, social
media, celebrities and other public figures
• Different types of mis/disinformation
• Conspiracy theories
• Misreporting (e.g. of death rates)
• False Tips for Prevention/Treatment/Recovery
• Scams (e.g. miraculous products)
3. Popular Cases
• Origin (WIV leak, bio-engineered, Pirbright
Institute, Bill Gates, etc.)
• 5G mobile phone networks
• Nurse whistleblower videos
• Ethnicity-based resistance/susceptibility
• Religious protection
• False treatments (alcohol, herbs, cocaine, …)
4. The Case of Masks
• A topic that caused a lot
of debate
• Asian countries clearly in
favour of their use
• US initially dismissed the
use of masks
• The WHO started
recommending the use of
cloth face masks by the
general public very late
(June 5)
By James Gathany - CDC Public Health Image
library ID 11162, Public Domain,
https://commons.wikimedia.org/w/index.php?
curid=6701700
5. Misinfo on Contact Tracing
• For NHS app (when it was still a thing…):
• app will ask permission to access all of your contacts
• app users who walk past someone who later gets
‘flu’ symptoms will have to be quarantined
• iOS Exposure Notification API:
• “iOS 13.5 update installation will automatically
allow authorities to track phone users’ locations and
who they’ve met with”
• “Bill Gates Negotiated $100 Billion Contact
Tracing Deal With Democratic Congressman
Sponsor of Bill Six Months before Coronavirus
Pandemic”
6. Misinfo Impact on Contact Tracing
App Success
• Contact Tracing App Success largely depends
on the adoption of the app by citizens
• Citizens hesitate to install such an app:
• valid reasons (e.g. apps are not private-by-
design)
• due to misinformation (e.g. false beliefs that
apps serve some malicious purpose)
• Ultimately, misinformation gradually erodes
trust and this translates to lower adoption
rates
8. Key Question
• Is it reasonable to expect that CT apps will be
installed by a large enough portion of the
population?
• Target: 56% of population, 80% of smartphone
users (Hinch et al., 2020) [does not mean that
smaller rates don’t have any effect]
• Evidence: 74.8% on a sample of 6000 adults in the
US, UK, Germany, Italy and France (Altmann et al.,
2020), but a US study by Kreps et al., 2020 on 2000
Americans revealed much lower support (30-40%).
• Real-world installation rates are promising but far
below the above targets
9. Key Conclusions
• Privacy and trust are the primary factors that
seem to determine adoption rates
• While privacy can (in theory) be ensured by
following best practices, trust is more elusive
• Trust “hygiene” factors:
• Transparency in communications
• Evidence-driven and fair decision making
• Availability of accurate and accessible information
• Countering of misinformation narratives
10. Thank you!
Get in touch:
Symeon Papadopoulos / papadop@iti.gr / @sympap
MeVer team: http://mever.iti.gr/ / @meverteam
https://virtual.globalfact7.com/
Editor's Notes
Hello everyone. I’d like to offer a perspective on the discussion with a different starting point. The majority of debates I’ve seen around the topic have mostly focused on the key questions of effectiveness and privacy. A related but slightly less discussed aspect of contact tracing that I’d like to bring to the table has to do with the issue of disinformation and how this affects trust.
The complexity and pervasiveness of the pandemic had a massive impact on our lives. It also dominated the public agenda; this led to a flood of information at unprecedented scale, variety and velocity. Much of this information turned out to be false, inaccurate or even fabricated. Misinformation found fertile ground and spread through a complex ecosystem comprising established media, social media, public authorities, politicians, celebrities and other public figures, and it was probably the first time where we witnessed such a diverse set of actors getting involved in the spread of various types of misinformation, which includes among others conspiracy theories, misreporting, false tips for prevention/treatment/recovery and scams. In parallel to the actual pandemic, an Infodemic was set in motion.
Several pieces of misinformation became widespread. Many of them pertained to the origin of the virus. For instance, several sources claimed that the virus was the result of a leak from the Wuhan Institute of Virology, others that it was bio-engineered, while others attributed it to a plan involving the Pirbright Institute or a population control scheme designed by Bill Gates. Similarly, several people blamed the 5G mobile phone networks, while others doubted the official statistics provided by governments backing up their claims with fake nurse whistleblower videos. Others saw an opportunity for promoting racial, ethnic and religious hate, and unfortunately many fell for a number of false treatments and cures that turned out to be ineffective or in the worst case lethal.
While for many of us such cases seem easy to debunk or ignore, we shouldn’t underestimate the extent to which they are adopted by different parts of the population, for instance marginalised groups.
In contrast to the previous clear-cut cases of misinformation, one notable controversial case where a lot of conflicting information was provided by multiple official sources was the public use of face masks. Although authorities especially Asian ones have been recommending the use of face masks in public, in many other parts of the world, conflicting advice caused confusion among the general population. Several governments and institutions, such as in the United States, have initially dismissed the use of face masks by the general population, often with misleading or incomplete information about the usefulness of masks.
It was only in early June that the WHO started recommending the use of cloth face masks by the general public.
This was also backed by a meta-analysis of 172 coronavirus-related studies which concluded that face mask use accounted for a large reduction in risk of infection.
This illustrates the complexity of the questions arising in such a volatile environment and that the notion of truth is constantly evolving.
And how is Contact Tracing related? For starters, there have been a number of misinformation cases about contact tracing. For instance, when the NHS app was first announced by the UK government, several false privacy concerns became very popular points of discussion in certain Facebook Groups. Also, a popular misconception pertained to the iOS Exposure Notification API and that its installation would supposedly allow authorities to track phone users’ locations and encounters. An even more far-fetched theory implicated Bill Gates in the negotiation of a $100 Billion Contact Tracing Deal with a Democratic Congressman Sponsor of Bill six months before the coronavirus pandemic.
https://twitter.com/Thomas1774Paine/status/1274865948206551040
Sensational thread: https://twitter.com/RVAwonk/status/1274386951144169472
But the actual impact of misinformation on contact tracing app success is potentially bigger. Misinformation gradually erodes trust in authorities and is therefore expected to lead to increased citizen concerns about the intentions of having such apps installed in as large part of the population as possible. It is hard to imagine that a person who believes in 5G and population control conspiracy theories would be friendly to the concept of a state-provided app for contact tracing.
The 2020 Digital News Report by the Reuters Institute gives some interesting insights into the different sources that people trust in terms of news. Fortunately, scientists and doctors top the list at 83% rates, followed by National Health Organizations at 76% and news organizations at 59%. More alarming is the fact that people have considerably less trust in politicians (only 35%). While big social media and search engine companies are not also perceived as trustworthy.
Given the above, a key question is the following:
Is it reasonable to expect that Contact Tracing apps will be installed by a large enough portion of the population?
A well-known simulation model by epidemiologists at the University of Oxford indicated that at 56% population penetration, these apps would really have a positive effect. A user survey conducted by an international team of researchers in Europe has indicated that user acceptance of such apps is quite high, so we should be optimistic. In contrast, a US survey carried out by Cornell University and MIT researchers led to a more pessimistic view measuring an acceptance rate between 30 and 40% among Americans.
It is worth noting that current installation rates of contact tracing apps in Germany, Italy and Norway are promising but still far below the targets.
Hinch, R., Probert, W., Nurtay, A., Kendall, M., Wymant, C., Hall, M., & Fraser, C. (2020). Effective configurations of a digital contact tracing app: A report to NHSX. en. In:(Apr. 2020).
Altmann, S., Milsom, L., Zillessen, H., Blasone, R., Gerdon, F., Bach, R., ... & Abeler, J. (2020). Acceptability of app-based contact tracing for COVID-19: Cross-country survey evidence. Available at SSRN 3590505.
Contact-tracing apps face serious adoption obstacles, https://www.brookings.edu/techstream/contact-tracing-apps-face-serious-adoption-obstacles/
To sum up, I feel that privacy and trust are the primary factors that seem to determine adoption rates of contact tracing solutions, and ultimately their success. While privacy has rightfully dominated the discussion around contact tracing app design and implementation, I believe that we should also look carefully into how we can integrate trust as a key component of the contact tracing app deployment strategy. As a starting point for such a discussion, I believe that the following trust “hygience” factors should be considered:
Transparency in communications about the app
Evidence-driven and inclusive decision making
Availability of accurate and accessible information
Countering of misinformation narratives.
Thank you very much for your attention!
For those interested, there’s virtual conference on fact-checking, which contains very interesting panels and presentations.
I’ll start with a cartoon that summarizes well the state of online information and disinformation. Once we became aware of the first COVID cases in China, there were plenty of us (me included) that potentially underestimated or didn’t pay enough attention to the problem. It was only after an increasing number of cases started appearing in Europe that we started paying more attention. This sums up very well that in hindsight many of us can get a clear picture of a situation, but when we are in the middle of the problem, reality gets a bit blurry.