I knew of Rafiq Copeland long before we ever met. I was managing Meta’s global information ecosystem program, and he was working to improve social media for some of the world’s most vulnerable and underserved people. Those displaced from their homes, affected by crisis or conflict, or who speak languages that platforms don’t often support. For these communities, having access to reliable information can be a matter of life or death.
Rafiq and I sat down to talk about the people that experts and institutions serve, and why human connection remains at the heart of building trust. Answers are edited for length and clarity.
◾
Jena Wuu: Hi Rafiq! So tell me, what is an information integrity expert? How did you get into this work?
Rafiq Copeland: I got into it in a roundabout kind of way. I was already doing humanitarian work in the Dadaab refugee camp in Kenya. It was 2011, and there was a famine in the Horn of Africa. The camp was expanding very rapidly with thousands of people coming in every day.
Internews was interested in building a radio station at the camp, and I thought it was such an incredible idea that filled a huge need. The camp had half a million people in it, with no formal communication channels of any kind. That created huge problems with rumors, which led to security problems. There was also a lack of accountability because refugees weren’t able to express any issues themselves.
I had worked in community radio throughout my studies. I was actually into podcasting before podcasting became a thing. So when I heard about this idea to build a radio station, I said yeah, I think you should do it, and then you should hire me to come and help run it. And that’s what happened. The radio station is still going, which is fantastic. The camp’s still there, which is not as good.
Ever since then, I’ve been working in that space: the intersection between information and accountability, mostly within crisis-affected communities or in contexts where accessing and sharing information is challenging, due to authoritarianism or something else.
JW: How does it work when you’re building communications in different cultures and languages that you don’t necessarily speak?
That’s a good question. I lived in South Sudan, and then I was in Myanmar for four and a half years. And I’ve worked in other places as well for shorter periods of time.
It seems crazy to be there advocating for the importance of meaningful communication when I don’t speak the language myself. That’s more common than you might think. In almost every case I’ve worked, at least one of the key parties shaping the flow of information comes from outside the local context. For example, in humanitarian response, that’s the UN agencies and NGOs providing aid and services to people affected by crisis. And in the last 10 years of my career, it’s been tech companies that operate online platforms.
So a lot of it is really about trusting the people on the ground experiencing events, and being an intermediary between them and those in a position to provide information. The people that should be more accountable to the communities they’re serving. My role is really to try and get those people in the room, and then step back and let them have their stake.
JW: That’s a pragmatic way to go about it. And as trust shifts from centralized authorities to individuals, it’s really going to be about community organizing and empowering people to reach others.
RC: I think that’s right. For our first broadcast at the radio station, we needed someone who could answer some questions from the community, and we needed a Somali speaker. There was no one from the UN agency responsible for the camps who spoke Somali and was in a high enough position that they would be allowed to come on the radio. So there were half a million Somali speakers in the camp, and no one who was authorized to speak to them. It was able to be rectified, but that was an embarrassing situation.
It’s been 15 years since then, and there’s still a lot of work to do. But there has been a lot of progress around this kind of engagement with affected communities.
A lot of it is really about trusting the people on the ground experiencing events, and being an intermediary between them and those in a position to provide information. The people that should be more accountable to the communities they're serving.
JW: We first met while I was working at Meta. From my perspective, Internews was a very important partner to the company – but I’m curious to hear about your experience of it.
RC: My first real engagement with tech platforms was after the Rohingya genocide in 2017. By now, it’s well-known that Facebook played a role in inciting the violence there. Even more so, the platform was being manipulated to create an enabling environment for a really toxic situation.
That had been documented for years by local civil society representatives. It became my role to help Facebook engage with those local experts, who really understood the context in a deep and meaningful way.
By early 2018, the Rohingya genocide became an international story. When Mark Zuckerberg testified to Congress that year, I think Myanmar was the only country besides the US that he was asked about in the hearing.
And all of a sudden, there was a lot of interest from Facebook to engage with the issues. And a lot of really smart people were hired.
JW: Yes, press is always an effective motivator. Some of my closest teammates joined Meta (then known as Facebook) around that time, who specifically had humanitarian backgrounds.
RC: It became evident to me that part of the reason why situations like Myanmar were allowed to happen was that the companies just didn’t have a clear understanding of what was going on, and what their own role was in it.
That’s not to minimize their responsibility. In many cases, they’d been told about it. But still, they didn’t really understand.
That was true in many complicated contexts all over the world. The situation in Myanmar was being replicated elsewhere, and there were lots of cases where we were trying to facilitate engagement and help the platforms understand what was going on, so they could do better to keep people safe.
JW: What changes did you want Facebook to make?
RC: There were certainly problems with content policy, which were discussed at great length. But in many cases, the policy isn’t the problem; it’s enforcement. We can talk about the policy all you like, but if you don’t enforce it, then the policy doesn’t really matter.
Enforcement was difficult for a whole range of reasons. In many cases, the problem was language capacity. When you have low language capacity, automated detection doesn’t work so well. So you’re reliant on users to report policy violations. And in many cultures, people just don’t feel comfortable reporting stuff. They might be concerned about getting in trouble – that the person they’re reporting might find out. Of course, the more authoritarian the country, the more likely that people feel nervous about reporting.
And in many cases, Facebook’s reporting features weren’t even localized into the local language. So unless you spoke English, you functionally could not report something, even if you wanted to.
Essentially, you’d wind up with an unmoderated platform because the automated detection doesn’t work, and people weren’t able to report issues either. As a result, nothing was getting flagged for human moderators to review.
Early on in Myanmar, I think Facebook only had two reviewers for the whole country. There was a lot of criticism about that, but that was actually an appropriate number for the amount of content getting flagged for review. Addressing the real issues was much more complicated than just hiring more Burmese speaking moderators.
JW: That’s one of the challenges with these huge global platforms that are built on very Western-centric, even San Francisco-centric culture. When I worked at Uber, we relied on trip ratings because there was no other way for us to know when there were issues. But people are scared to report, or to get someone else in trouble. It’s a lot of responsibility to put on users.
RC: And then in other cases, you can have the opposite problem, where people massively over-report things that they just don’t like or things that they disagree with. So how do you calibrate all of that?
JW: So true, I definitely saw that at Facebook. Just wondering… did you end up with a favorite platform based on your experiences working with them?
RC: I had a real love-hate relationship with Facebook. It was where we saw the most negative impact at that stage, although YouTube gives them a real run for their money.
But Facebook was also the platform that engaged the most with organizations. They really did far more than any of the other companies to address the problems. I mean, the bar is low. But I met and worked closely with a lot of people at Facebook that were really doing their best to make positive change.
That was something that surprised me. The companies seem so monolithic, but then you find there are people within the companies that are really fighting internal fights and investing a lot of their own political capital. One of the most effective ways to work together is finding those people and giving them the material they need to have those internal conversations.
When you have low language capacity, automated detection doesn’t work so well. So you’re reliant on users to report policy violations. And in many cultures, people just don't feel comfortable reporting stuff … Essentially, you’d wind up with an unmoderated platform.
JW: There is rightfully so much scrutiny on the big platforms. But the reality is people communicate in other ways too, which are just as important for experts to engage in. What are other information flows that people rely on?
RC: The one that gets overlooked all the time is people just talking to each other in real life. When you talk about trust, those real life interactions still massively outweigh online communication.
If you’re trying to impart information, then the environment you most want to be in is the conversation at places like the nail salon. That’s why influencers try to build rapport, right? It’s recreating those human connections in the online space.
We also need to remember that the big platforms aren’t a monolith. I’ve done surveys with people where they say that Facebook is the primary place where they get information. But whereabouts on Facebook?
It’s almost always very local pages that are set up for something else. For example, I had a case where the local car dealership was posting about whatever was going on in town. Weddings, or car accidents, that kind of thing. And people would tell us, he’s one of the people they get their information from. So it’s on Facebook, but it’s still about that human connection.
JW: Definitely. Do you see that evolving as technology changes?
RC: A lot of conversations are happening in messaging groups now. I’m sure stuff like the car dealership posts are all on WhatsApp groups now.
And that’s a different dynamic. It’s certainly harder to study because it’s less public. But the interesting thing about those spaces is that they are very intimate. It may be the chat group for your street, or your family chat group. But as they get bigger, they become effectively anonymous. So you have this shift between something that’s very personal, and something that’s the complete opposite of that, and it can shift quickly between the two.
[People] say that Facebook is the primary place where they get information. But whereabouts on Facebook? It’s almost always very local pages that are set up for something else … So it’s on Facebook, but it’s still about that human connection.
JW: If we know that some of the highest trust spaces are these small chat groups, what can organizations do to engage with them? Because they’re essentially private.
RC: This is really hard, especially in the metrics-based funding environments that we’re in. If your job as a communications person at an NGO is to get out some messaging, you can count impressions on a post. But if you’re really trying to inform people in those private spaces, it’s out of your control because you’re not the one sharing directly. That becomes really difficult to measure, and I wish we could tolerate that more than we do.
Obviously, there are things you can do to measure impact. You can do research, and talk to people. But then it takes more work and becomes more expensive just for measurement.
JW: And when it comes to mitigating crises, it can be a double negative to have to prove that your work prevented something bad from happening. But when chaos does happen, it’s a major issue.
The migrant crisis in Ceuta is a recent example, where a lot of people were making risky, life-changing decisions based on information being shared in chat groups. How can organizations prepare for helping with those information flows?
RC: Migration is a really interesting example because there’s been really good work in that space, but it’s often quite controversial.
I was in Greece in 2016 during the European migration crisis around the Syria conflict. We had an online information service that was updated every day. It had news and information about visa processes, when borders closed, things like that. It was the sort of information that’s changing all the time, but that you would need to make good decisions and not put yourself at risk. But there’s often a lot of criticism for that kind of thing.
The important thing is: you can’t just provide a sanitized version of the information because then people won’t trust it. If a border closes, you should tell people so they won’t try to cross it. But if the border re-opens, you have to tell people. Because if they see you’re withholding, then they’ll stop coming to you for information. So you need to be transparent.
JW: I can see that on one hand, you want to help people making serious life decisions. And on the other, even sharing factual information can be seen as a political or moral position, depending on what the facts are that you’re trying to put out there.
RC: This is the problem with being an actor in a contested space. You’re not telling people any secrets. You’re just making it easier for them to find information that already exists. But in a controversial topic like migration, it can feel like you’re taking a stance just by getting involved.
You can’t just provide a sanitized version of the information because then people won’t trust it … if they see you’re withholding, then they’ll stop coming to you for information.
JW: That makes me wonder: Is there a space for humanitarians and civil society organizations to share more publicly about their work, so that people understand what they’re actually doing?
RC: Yes, but it’s hard and complicated. When you don’t talk about the work and why it’s important, you cede ground to the counter-narrative, which is largely incorrect.
That said, there are reasons why some organizations might need to take a more cautious approach. Ultimately, organizations have a responsibility to the people that they serve. And that leads to quite a conservative approach when it comes to communicating with the general public.
JW: Is there anything that we, as the general public, can do to help support experts who are taking the risk of talking about their work online in these difficult situations?
RC: Hm… like and subscribe?! In seriousness, it’s about giving your support through engagement and expressing support more publicly. In a lot of cases, one side of the conversation is very noisy and the other side is very quiet. It’s easy to assume that one side’s bigger than the other, and that’s not always the case.
People talk about virtue signaling as if it’s some awful thing, but there are worse things to be signaled than virtue.
◾
Thank you, Rafiq Copeland, for sharing your expertise! You can find Rafiq on LinkedIn. ◾