A new Twitter policy cripples journalists’ efforts to halt disinformation

The 2024 election season is upon us, and experts are already predicting that advances in technology could supercharge the spread of disinformation over the coming year and a half. Purveyors of disinformation are poised to severely undermine election integrity and erode confidence in the electoral process. 

More than ever, the public will rely on journalists to detect and expose this disinformation in their reporting. But, under Elon Musk’s leadership — which, ironically, began with a focus on eliminating bots on the platform —  Twitter’s newly amended application programming interface (API) policy may rob journalists of access to bot detection tools, which are critical to identifying and understanding the spread of disinformation on social media.

Tools such as Bot Sentinel, Botometer and Hoaxy access Twitter’s API to read and analyze inauthentic or automated platform activity. Bot Sentinel and Botometer use this data to determine which accounts exhibit bot-like behavior and then provide classifications and ratings of accounts. When stories are shared on social media, journalists can use Botometer and Bot Sentinel to assess the accuracy and legitimacy of these stories and their sources. They can then turn to a tool like Hoaxy to visualize the spread of information from bot-like accounts and identify where mis- and disinformation may be influencing stories shared online. 

Through these tools, journalists can avoid inadvertently passing on disinformation in their reporting and can expose and debunk false narratives. 

These resources are essential in today’s fast-paced and often under-resourced news environment. In a nationwide survey by PEN America (where I am employed), only 14 percent of journalists reported that their newsrooms had a dedicated in-house fact-checking team to monitor and debunk disinformation. Bot detection tools can be a game-changer for exposing targeted falsehoods and conspiracy theories, especially for small, local newsrooms serving marginalized communities.

Until recently, the API access necessary to operate these tools was available for free or at a low cost. But in March of 2023, after Twitter implemented its new API access policy, the price to operate these tools went from free (or nearly so) to potentially $42,000 a month for access to the same, or even less, data than was previously available.

Needless to say, these changes will likely make it impossible for academics and nonprofits, such as the Observatory on Social Media (OSoMe) at Indiana University — which operates Hoaxy and Botometer — to continue to provide these vital tools. Even for those that can afford the $100 monthly Basic API subscription, the access it provides is woefully insufficient to maintain the previous usefulness of these tools — just 0.3 percent of what used to be available for free in a single day, according to the Coalition for Independent Technology Research.

The disastrous fallout from these changes has already begun. Hoaxy no longer has access to information necessary to operate and is currently unavailable. Botometer expects to stop working in the coming days. OSoMe has stated that it is “actively seeking solutions” to keep these tools operational but warned that even if it can build new versions of the tools that will work with amended API structure, they may have “limited functionalities and quotas” due to Twitter’s access restrictions. And although Bot Sentinel’s founder Christopher Bouzy has vowed that “Bot Sentinel is here to stay” (the classifier tool is once again operational after being temporarily offline), it remains to be seen whether its functionality will be impacted.

Forcing researchers and nonprofits to pay exorbitant fees to view a fraction of the data previously available to them for free is outrageous and threatens to strike a devastating blow to journalists’ ability to combat the spread of disinformation.

Twitter can, and should, roll back these crippling, unjustified fees. At a minimum, it should make API access free for public interest uses across the board — including for bot detection tools and tools that help protect journalists against online harassment — as it did for government and publicly owned services that use automated bots to tweet public alerts. 

As we enter what is likely to be yet another turbulent election season, journalists need all the support they can get to fight back against distortions and outright lies. Twitter shouldn’t be disarming them just as the fight gets started.

Shannon Jankowski is the program director for journalism and disinformation at PEN America, the free expression and free speech organization.