Twitter announced in April that access to the API will now require a paid subscription, with those most useful to researchers ranging from $42,000 to $210,000 per month. The change has left many policy shops, NGOs, independent researchers and students without access.
“It’s a major concern,” said one international human rights investigator, who spoke on the condition of anonymity because they were not authorized by their organization to publicly discuss the matter. “A lot of major institutions, ourselves included, could struggle to budget for that.”
Musk has moved to begin charging for several services that Twitter once offered for free, including Twitter’s signature blue check mark badges, which previously denoted that accounts at risk of impersonation had been verified as authentic. He has justified charging $8 per month for the badges in part by arguing they would reduce the swarms of bots.
“Verified accounts are 1000X harder to game by bot & troll armies,” he said in a tweet in March. He has argued that scammers would be unlikely to pay large sums for the ability to spam users.
Although Musk has repeatedly claimed that his efforts have been effective, there has been no discernible drop in bot traffic since the billionaire mogul bought Twitter for $44 billion in April of last year, according to the cybersecurity firm Cheq, which monitors traffic on the site.
Musk’s takeover sparked a “reprioritization of what Twitter means to the world,” said Alexa Koenig, the co-director of the Berkeley School of Law’s Human Rights Center, which has worked with the United Nations to codify international standards for conducting online research of alleged violations of criminal, human rights, and humanitarian law.
“I think right now, any responsibility to respect, protect and remedy human rights violations that the company may be causing or contributing to is very clearly taking a back seat to the financial interests,” she said.
Charging for check marks has been part of an aggressive monetization strategy that Musk has employed to generate revenue — in part to pay back the debt he took on when he bought the site.
“We need to pay the bills somehow! Twitter cannot rely entirely on advertisers. How about $8?” he wrote in a tweet to author Stephen King in October, shortly after taking over the site.
The API change has sparked concern from some lawmakers in Washington.
“This move will make it more difficult for researchers to access the information necessary to understand harms on Twitter,” Rep. Lori Trahan (D-Mass.) said in February.
Twitter has shed about 80% of its workforce since Musk’s takeover, including its human rights division and its Moderation Research Consortium, which investigated state-backed attacks on the platform. The company did not respond to a request for comment.
Although other social media platforms, including Facebook, have given researchers free access to their API in the past, Twitter stood apart because of its relative transparency. Company liaison officers helped prosecution teams hunt down information. Open source investigators were given access to huge public data sets, allowing them to create tools to organize evidence of potential war crimes and establish a chain of custody that would be permissible in court.
Human rights investigators focused on Syria and Ukraine said that Twitter had been a crucial tool in locating disinformation clusters around wartime incidents that Moscow sought to blame on its opponents, including chemical weapons attacks by the Syrian government and Russian strikes on Ukrainian schools. Evidence gathered from Twitter of Russian war crimes in Ukraine has even been submitted to the International Criminal Court.
In Syria, where droves of civilians were being killed by government airstrikes, one organization, Hala Systems, used the API as a predictive input for an early warning system that helped rescue workers prepare as warplanes approached.
“Being able to capture potentially relevant activity either takes an army of people working shifts in a basement somewhere, which we do not have; or some relatively smart computer programmes, which we do have,”said John Jaeger, a former U.S. diplomat who co-founded the company.
“This wasn’t something that was conceivably taxing Twitter infrastructure and I would be surprised if it was usage that cost them a meaningful amount of money,” Jaeger said.
“They’re taking away a tool that is broadly useful to society, particularly those working to support a fact-based reality and accountability for atrocities,” he said. “Not by taking it away in its entirety, not by turning it off or saying it violates terms of use, but by simply putting it out of the reach of organisations like ours and certainly of dedicated individuals.”
Since changes to the API policy will also limit the number of tweets that an account can publish without access to the enterprise tier to 3,000 per month, Jaeger worries that Hala will no longer be able to use Twitter to issue warnings in the event of an intense aerial bombardment by the Syrian army.
During the spring of 2018, airstrikes killed and wounded more than 5,000 people in just two weeks.
Twitter’s free API policy also facilitated the discovery of some of the most influential bot campaigns to date. Election meddling in the United States and several African countries by the Kremlin-backed Internet Research Agency would have been much harder to expose without it, researchers say, as would the same troll factory’s disinformation campaigns in the Middle East and Africa.
In the early months of Sudan’s bitter conflict this year, researchers at the Atlantic Council’s Digital Forensic Research Lab used the API to reveal how hijacked Twitter accounts were used to artificially amplify the narrative of one of the warring sides, and to boost English-language posts that might appeal to an international audience.
Next year, national elections are scheduled in Chad, Mali, Rwanda, Somaliland, South Sudan, Pakistan, Sri Lanka, Tunisia and Venezuela. In a recent report on the API changes, researchers with the International Crisis Group concluded that “this policy will significantly reduce information about the impact of election-meddling campaigns, the online harassment of activists and effects of disinformation on violence in countries where rule of law is fragile and independent media outlets do not or have ceased to exist.”
After an initial backlash to the API changes, Twitter announced in May that verified government or publicly owned services that tweet weather alerts, transportation updates or emergency notifications could still use the system for free. “One of the most important use cases for the Twitter API has always been public utility,” the platform said.
But developers and researchers say government-linked exemptions aren’t enough.
In the days after massive, deadly earthquakes struck Turkey and Syria in February, killing almost 70,000 people, independent software developers rallied together to create databases and heat maps of where people were trapped in the rubble as government rescue efforts lagged.
“In a developing country, if an earthquake or a flood or any other disaster of this magnitude were to take place, current API rules could significantly impact rescue efforts,” said Akin Unver, a professor of international relations at Ozyegin University in Istanbul who worked on the mapping effort.
Alessandro Accorsi, who led the Crisis Group research, said that many humanitarian organizations likely couldn’t afford the new API rates.
“It means they are not going to be able to use it, and in an emergency situation or in a conflict, especially where there is weak media or major disruption of infrastructure, its information is vital. It saves lives.”
Faiz Siddiqui in San Francisco contributed to this report
