I predict we’ll see more social media posts in news stories — but still without the necessary verification of who their authors really are.

Social media posts increasingly make their way into news stories as sources — a new vox pop — representing public opinion on everything from celebrity drama to presidential debate performances. A recent study suggests that the use of social media content in the news has nearly doubled over the past five years. These posts come overwhelmingly from Twitter, even though less than a quarter of U.S. adults use the platform.

Including social media posts in stories as representations of public opinion offers at least the possibility to cast the American public as the protagonist in news stories, as Danna Young called for in her prediction last year. Unfortunately, my research suggests that for now, social media posts are mostly used in political news to further propagate the game frame.

Previous predictions have called for an increase in automated fact-checking, but we also need to automate social media verification. Unfortunately, we’re nowhere near that. The first step is to integrate social media verification at all.

I’ve interviewed journalists about their use of social media posts from the public in news stories. Journalists told me they select posts based on alerts from DataMinr, by following hashtags, or even after seeing them pop up on their own Twitter timelines. None of them mentioned verifying the authenticity of the post before incorporating it into their reporting. Given the outsized attention that some journalists pay to Twitter when determining what’s newsworthy, I see little reason to think that the use of tweets in news stories will decline in 2019.

I’m not contending that journalists skip verification on social media posts out of laziness. My suspicion is that journalists think of tweets as content, not as sources.

When journalists cover tweets from the president — as they often do — they almost always fact-check the claims Trump makes in his tweets. The same must go for social media posts from the public. The authenticity of their account must be verified, and the veracity of their claim (if it’s not straight opinion) must be established.

Verification would also work to mitigate the risk of news outlets embedding social media posts from bots, trolls, or other streams of manipulation into their stories. Tweets from what are now known Russian troll accounts found their way into news stories in major outlets like The Washington Post, USA Today, BuzzFeed, and Vox. This is an opening that can be exploited, as Nick Diakopoulos pointed out in his prediction last year.

Since embedding tweets straight from Russian propaganda accounts didn’t seem to prompt newsrooms to institute verification procedures, I predict 2019 will bring more news stories that unintentionally amplify malicious misinformation. At a time when the press and journalists are under attack and trust in news is at an all-time low, it should go without saying that these verification steps are vital.

We obviously can’t rely on platforms like Facebook or Twitter to validate the authenticity of accounts, nor can journalists expect to rely on them for data to do so on their own. As a first step, journalists should incorporate standard verification procedures into the social media reporting process by at least attempting to contact users, verify their authenticity, and perhaps even seek permission to use their posts in news stories.

Once journalists begin to do this, we can learn what the process of social media verification looks like. Only then can productive partnerships between news outlets, journalists, journalism researchers, and computer scientists be forged to address the problem of social media verification at scale. I hope that if my prediction comes true, it will shepherd in much-needed collaboration on solving the social media verification problem.