X’s misinformation woes get worse during the Israel-Hamas conflict

Last weekend, when Hamas launched a surprise attack on Israel, people around the world flocked to Twitter — now X — for up-to-the-minute information.

What they found was a site crawling with misinformation: images captured months or years earlier in unrelated attacks, inaccurate claims about other countries entering the conflict, even a fake White House press release announcing billions of dollars in new U.S. aid to Israel made the rounds.

And X’s new owner, Elon Musk, promoting accounts known for spreading lies and hate didn’t help.

The signal-to-noise ratio on X is worse than ever, said David Clinch, a founding partner of the social media intelligence agency Storyful and co-founder of Media Growth Partners.

Marketplace’s Lily Jamali spoke with Clinch about what X users should remember when scrolling through the platform for news on the Israel-Hamas situation. The following is an edited transcript of their conversation

David Clinch: Don’t trust anyone online, including yourself. And what I mean by that is, you know, somebody in a breaking news situation, your urge is to see an image or a video that’s described as showing something and that it’s just happened, your urge is to share that immediately. Don’t trust your own urge to do that. Stop a minute, think about who posted that image or video. You know, can you see other people, more reliable people, posting it or sharing it? Are there comments below it which tells you whether it’s real or new or not new? So there are relatively very easy, quick ways to double check. And if you’re not sure, don’t share. And I tend to avoid that. I sort of have a curated list of reliable sources, so I’m very rarely seeing any of those reliable sources sharing something that’s old or misleading. It does happen. People can make mistakes. Journalists or others on the ground can make mistakes or be misled. And those images and videos, of course, are becoming really sophisticated, that some are not just old and misleading, they may even have been AI generated. But the most important thing is that you should not share an image unless you’re absolutely certain that it is real. That’s an important thing for most users of any platform to think about when they’re trying to follow news.

Lily Jamali: And [that] was a best practice, you know, [and] has been a best practice for a long time now.

Clinch: Yeah, absolutely. I mean, back in my Storyful days, Storyful was and is an agency for forensically verifying images and video that emerged during breaking news. And I think the best lesson is don’t share unless you’re sure. But also, there are some ways to think about verifying yourself. You can do Google image reverse search. And again, you can look in comments to see whether people are affirming that the video is real or not.

Jamali: Well, what’s interesting too is that X recently started paying certain content creators, which has led to backlash about what that’s done to the incentive structure. I mean, let’s be honest, there’s always been this issue about there’s an incentive to post incendiary comments, or, you know, things that will have a lot of engagement. But it seems like paying people to do this is only making that issue worse. Is that your read on it?

Clinch: Absolutely. It’s bad incentives. It’s basically Reply Guys meet revenue. And if you put those two things together, it’s just an awful dynamic, because you’ve got people whose only interest is clicks and followers. And then you’ve given them the gift that these clicks and followers help them make money behind the scenes. That is absolutely [an] upside-down set of incentives. And I think that the leadership at X, if they’re interested in having a reliable platform, they should change that. And they should switch those incentives back to reliable journalism, to reliable information. It doesn’t have to be an official journalism source. It can be verified, it can be forensically proven to be real, and then it can still be valid, and they can still generate interest and even money in some cases. But it shouldn’t be the incentive, [it] shouldn’t be that just people with lots of followers get to share whatever they want and make money behind the scenes.

Jamali: One opinion writer over at Bloomberg said it’s turned the site into a “hive of engagement farming.” Another thing I wanted to ask you about was at one point over the weekend, we saw Elon Musk, who of course now owns X, referring people to a user known for spreading misinformation and another who has shared antisemitic views on the platform, calling them good sources for real-time information. Do you have any thoughts on that?

Clinch: I do. I think that is incredibly dangerous. Elon Musk has a lot of followers on this platform that he now owns. He should know better than to share sources that aren’t — I mean, apart from anything else, they’re not real sources that are on the ground actually witnessing what’s going on. Even if theoretically there are people who are good sources outside of the region, these sources definitely aren’t there on the ground. And then even more than that, a quick check would show that not only are they unreliable, but they are incredibly biased and dangerous. That’s dangerous. The tragic thing is that there are reliable sources. I know he talks about citizen journalism. And I agree, there is a huge role for eyewitnesses who are there, who are closer to the story. That was our mantra at Storyful: There’s always somebody closer to the story. But you have to verify that they’re actually closer to the story. You have to put context around the fact that they’re not professional journalists. You can’t assume that they’re showing you everything and telling you everything. There is verification that’s required.

Jamali: So, David, based on what you saw this weekend, how has the role that the platform plays in real-time conflicts changed?

Clinch: I think that there is still huge value in information that emerges in real time on Twitter. There are certain things that have happened in Israel and Gaza over the last few days, videos that have emerged and things that I have been able to verify and see that have not been shown on television, have not been properly communicated by mainstream news. And I understand why, because it takes longer for those things to go through that vetting process. But nevertheless, there is an extremely important value to Twitter’s real-time ability to surface important information and video and images. The problem is that the negative sides that used to be sort of generic noise — and then you could get the news from the noise if you knew what you were doing — is now so much more magnified — the misinformation, the highlighting and amplification of unreliable sources and misleading information. And then, of course, the sort of vulnerability to AI or manipulation of that kind as well too. If you’re an average user, the noise is overwhelming. And the person who owns this is amplifying the noise, not the news.

Jamali: And what did you notice about how other people were trying to combat myths and disinformation on the platform this weekend? One thing I noticed, I think you actually retweeted it, was an account called Geoconfirmed saying they’ve opened up a map for users to post and share geolocations. How does something like that help keep users informed and help all of us who just feel like we’re in the dark here?

Clinch: Yeah. Well, I’m stunned. And I think the opposite of misinformation and the opposite of careless sharing is forensic verification. I am always interested in supporting anybody — the teams that do that, individuals that do that — that don’t just look at the piece of video and go, “Is it real, or is it not real?” But they want to know everything about it, because it can contribute to the story. So forensic verification, geolocation, confirmation of a video of the original attack inside Israel so that it’s not a rumor of what happened, that it’s forensically verified, that is actually an extremely valuable journalistic process that’s happening. And Twitter still plays a part in that, but that’s a higher skill set. That’s higher level, and something that I watch with admiration from afar, but it’s essential if you want to be responsible about covering news. You have to do forensic verification of every image, in every video, because no image and no video can capture everything, There’s more context, there’s more information. And that still gives me hope, because I see journalists doing that kind of thing every day.

More on this

An X user this weekend asked whether the company is still taking reports of misinformation and disinformation, noting that the option was gone for users wanting to report a specific post.

We reached out to X with that question and got no response to our emails.

A good place to start navigating news out of Israel and Gaza on X is David Clinch’s curated list of reliable sources posting on the conflict in real time.

You’ll also find a Washington Post article by journalist Joseph Menn, who interviewed a number of information researchers last weekend. He reports the consensus view among them is that X has become much less reliable since Musk took control of the platform almost a year ago.

And you can check out the Bloomberg Opinion piece we referenced in the interview on engagement farming. Columnist Dave Lee lays out his argument for halting engagement with Elon Musk’s X.

Correction (Oct. 10, 2023): A previous version of this episode misstated David Clinch’s title.