On February 2nd, @TwitterDev announced that as of February 9th, 2023, Twitter will no longer allow free access to the Twitter API (application programming interface).Twitter is justifying this change as a response to bad actors that have been abusing Twitter API access.
Twitter’s latest shift away from accountability will constrain how researchers, journalists and other groups understand complex social problems, including technology’s influence on our democracy. This change is concerning because it will place important data behind a paywall and hinder efforts to understand online toxicity, dis/misinformation and civic participation on a platform that has come to function as a digital public square. In a period of democratic backsliding, it is important that social media platforms mitigate bad actors through more transparency, not less.
The Twitter API allows people outside of Twitter to create programs and applications that can interact with the Twitter platform.
The Twitter API powers simple bots that tweet local information or perform functions like organizing threads or tweet reminders. It also supports more complex, large scale applications that provide insights on consumer trends or public health. The API serves as a crucial resource to track user behaviour and investigate potential threats to our democracy.
The February 2 announcement has received considerable backlash and in response Twitter has committed to allow free access for very simple API interactions. This means some basic Twitter bots may be here to stay but anything beyond simple usage will be constrained. Due to cost, many projects that are producing important insights based on Twitter data will not be able to continue.
Originally intended to come into effect on February 9, the end of free Twitter API access has been pushed to February 13. Subscription prices have not been officially announced, aside from $100 per month for “basic access,” which remains undefined. Rates for other levels are currently “TBD.”
At the Samara Centre we have used the Twitter API since 2021 for our SAMbot project, which tracks toxicity received by candidates on Twitter during Canadian elections. This data has helped to quantify the volume and intensity of abuse received by those on the digital campaign trail and provided important insight into the lived experience of those seeking public office. To date we have tracked Canadian elections at all levels of government and analysed over 3.7 million tweets, over 620,000 (or approximately 17%) of which were found likely to be toxic or abusive in nature.
SAMbot is one example of how Twitter data is used to uncover important insights about technology’s influence on our democracy. In the short-term we plan to continue this work, however, we are concerned about what this change to the API signals more broadly. Other important work in this field may be restricted or forced to end.
Previously Twitter provided greater data access than other online platforms distinguishing it as a key resource for the social media research community. A sudden and sweeping change to its API will have wide reaching consequences. This underscores how important it is to strengthen regulation and address the online harm and outsized reach facilitated, not just by Twitter, but by social media platforms more broadly.
Open data requirements foster informed research insights and better understanding of our society. This approach serves a public good and supports an ecosystem of ‘good actors’ that have shaped the platform in constructive ways. Twitter has an opportunity to demonstrate strength and relevance through increased transparency. ‘Bad actors’ always find a way no matter the cost. Instead of managing them, why not bolster the bots that make a difference?