On disinformation, the information ecosystem, and health security

I had an opportunity to on behalf of Dr Sylvie Briand recently visit Marcia Castro and the Department of Global Health and Population, Harvard T.H. Chan School of Public Health for a health security symposium for the 60th anniversary of the department’s establishment.

On a panel with Joan Donovan , Guilherme Canela de Souza Godoi, and Kevin Croke , we discussed disinformation, the information ecosystem, and how they impact health security.

Disinformation is inaccurate information that is specifically designed to achieve an agenda, and it is a component of an?#infodemic?during an emergency, alongside questions, concerns, information voids, and circulating narratives and?#misinformation.

What are some of the lessons from the COVID-19 pandemic regarding managing disinformation during emergencies?

• People who feel overwhelmed by information and are anxious about COVID-19 are more prone to acting on misinformation. In times of stress and crisis, we tend to believe the first thing we hear, have difficulty understanding complex information, and look for alternative sources of information.
• Disinformation only makes up a very small part of the sum total of conversations about COVID-19. Understanding larger narratives full of more innocent questions and concerns that go unaddressed, narratives that can be used to promote conspiracy theories, loss of trust in government.
• This is further complicated by people seeking information who can’t find it from credible sources, so they are more inclined to act on less credible information.
• Disinformation that is driven by financial motives is easier to uncover and address than those driven by political motives. They are also more amenable to fact-checking than politicized mis- and disinformation.
• Health authorities find it challenging to address health disinformation effectively because it ultimately politicizes the public health response. Therefore, partnerships and networks who are better able to credible address and mitigate mis- and disinformation become critical.

?

How has the spread of disinformation, especially in the past decade, affected global health security concerns?

• Public health is global health security. That clear relationship was firmly established during COVID-19, because now, all public health is global, and health security is not limited just to the borders of one’s country.
• The information environment we all live in is connected across borders and are no longer defined by communities that share common geography. That means targeted disinformation campaigns can cause major damage to all of those linked to that global information environment. This is why having infodemic managers being installed in more institutes of public health—like they have in Germany, the US, Indonesia, Nigeria, South Africa—just to name a few—will become more important. Being able to detect their narratives in “peacetime” and scale up in emergencies will be important for preparedness for tomorrow’s pandemics and emergencies.
• We have seen from Ebola, Zika, COVID-19, and mpox that mis- and disinformation gains traction in low-trust environments where people don’t trust their government, don’t trust health service delivery or the pharmaceutical and public health and social measures designed to protect them. Building resilience against disinformation means rebuilding that trust, but doing so in middle of a chaotic global emergency wouldn’t be something that would come easy or is recommended.
• We know that disinformation and misinformation becomes more dangerous if it advocates for people to not take emergency precautions, to ignore public health guidance, to advocate for stigmatization or violence against groups of people who are often marginalized already, such as refugees and migrants, and people who are vulnerable because of their work, such as health workers, or their health, such as pregnant women and the elderly. We need to assess relative levels of risk based on the populations involved and the context they are living in, because we can’t try and address all of the problematic and potentially harmful narrative simultaneously.
• Determining what are likely to be the most common narratives and disinformation-related narratives ahead of time can help us plan and develop that risk matrix that would also help us determine what kinds of events would trigger a “high risk” alert versus a “low risk” one, with different actions to match.

?

Are there best practices on introducing strategies to mitigate disinformation during the implementation of health policies?

• Ensure that credible, accurate health information is easy to find, easy to understand, and easy to share.
• Mis- and disinformation often uses emotionally manipulative content to accelerate its spread, which health authorities struggle to counter effectively with their own messaging. This means that addressing disinformation narratives may require out-of-the-box thinking.
• We can take a few lessons from some misinformation narratives that spread and how humor has been used to defang them. One prominent politician in Brazil said that COVID-19 vaccines turn you into alligators…so people showed up to their vaccination appointments with alligator costumes, and documented the experience on social media. In another example, in Indonesia, the misinformation that COVID-19 vaccines made you magnetic, “proven” by people putting coins on their vaccination spots that stick there. So a prominent social media influencer tried to stick a large denomination rupiah note on his upper arm, reporting he was irked that it didn’t stick and he wished the vaccine worked to attract higher value rupiah than just coins. This followed a flood of similar photo attempts to stick rupiah notes on newly vaccinated arms.
• Humor is a powerful weapon, but again, it may not be the health authority best positioned to use it…but someone in their network could.
• The bottom line is, if we don’t make health information as compelling as the mis- and disinformation, we will continue waging a losing battle, and that means investing in better content, better ways of sharing and amplifying it, and translating it to communities through new trusted messengers working in trusted spaces.

?

The body of literature around disinformation has expanded dramatically over the past few years – what information or analysis can we extract from the current disinformation research?

• People and institutions spreading disinformation have taken advantage of design flaws of social media platforms, such as recommendation algorithms for related videos suggesting extremist content.
• Content creators also get around content moderation attempts through a whole range of techniques, from as simple as using words other than “hot button” words that don’t get caught by automated text-based moderation or using written words held up on screen or even gestures to refer to concepts.
• Algorithms need to get smarter but so does human oversight, because memes and emotion and conveyed meaning are culturally specific and don’t translate well to automated detection.

How can health authorities contribute to protecting freedom of expression?

• We can start by establishing our own guardrails for how health institutions conduct social listening and Infodemic management in an ethical way, where governance and privacy regulations are adhered to and where the focus of analysis is population level and not individual level and protect the privacy of individuals who have a right to express whatever sentiments they have.
• We need to identify when those narratives get hijacked for sinister reasons and help people better critically appraise the information they received so that they can act on the best health advice to protect themselves and their families.
• Content takedowns, for example, are not a preferred method of addressing mis- or disinformation for health, because it will likely just pop up again, often on more extreme platforms. Instead, focusing on building communities that feel they can provide context, to fact-check claims, debunk conspiracy theories, and readily refer to high-quality information sources, can go a long way to make keeping a healthy information environment online not just a challenge for governments and tech platforms to solve.

How is public health challenged by the dynamics of the information ecosystem? How does the information ecosystem affect our trust in the health systems and our ability to promote health and well-being?

• I had mentioned before that low trust environments become breeding grounds for doubt and mis- and disinformation that can further erode trust in government. Often, health authorities and other trusted voices in health such as health workers and medical associations, are almost entirely absent from the online conversation.
• We can no longer assume that underinvesting in communicating in digital spaces is okay before the next pandemic.
• Ultimately distrust is the outgrowth of the perception that promises were broken and values were violated. That means that changing the information environment alone is not sufficient to regain trust. Health systems and response structures need to function well in routine and consistently in emergencies, in ways that are acceptable to communities, to rebuild trust.