Twitter: API change makes searches difficult – 02/10/2023 – Politics

After a back and forth, Twitter announced this Wednesday (8) changes to the policies of its system used by third-party programs to connect to the platform. If no one changes their minds again by then, the update will take effect on Monday (13).

The service in question is called an API (application programming interface). It is a common tool, mainly on large websites, and allows any programmer to create systems to interact with the platform.

With it, it is possible, for example, to extract mass posts, create robots to automatically share content or use the social network account to log in to other services without registering.

This is yet another controversial change implemented since Elon Musk bought the social network in October.

It can harm the work of researchers who rely on data collected on the platform, such as those focused on analyzing fake news and hate speech. In addition, the continuity of some useful robots on the platform is uncertain.

Today, virtually all of the tool’s features can be accessed for free, with limitations on the volume of data that can be manipulated.

There are also two paid options, one called “Premium” (costing from US$ 149 to 2,499 per month, or R$ 773 to R$ 12,964) and another “Enterprise” (without defined value), with more generous quotas.

Last Thursday (2), Twitter announced that it would extinguish the free version of its API from this Thursday (9). After the negative repercussions, the speech changed.

On Sunday (5), Elon Musk wrote that verified accounts could maintain access only to publish content – ​​there is no mention of this in the official announcement. Additionally, he posted that they would keep a free, simpler option for bots that provide “good content”.

The criteria or selection process for this is not clear.

Wednesday’s statement says the free version will be limited to creating 1,500 tweets per month per user, but doesn’t explain to developers how that will work. It also does not clarify who would have access to this modality, nor if it would allow continuing to extract content for analysis.

The announcement warns that the Premium version will be phased out and that its customers should migrate to the Enterprise model. A “basic” version, for access in small volumes, will be created for U$ 100 (R$ 519).

The change should not impact companies that consume a lot of the API so much, as these already have the highest paid access modality. Those who suffer are, for example, researchers who use the system to extract data to be analyzed.

“Here what concerns the most is uncertainty”, says Pedro Barciela, social network analyst. He is one of those responsible for “This is such a social network”, a site that publishes content evaluating the debates that take place on Twitter.

“This insecurity often ends up preventing some studies that would be carried out in the long term. For example, the analyzes carried out by US researchers on Russian influence in the country’s elections, which took months or years”, says Barciela.

Victor Piaia, a professor at FGV ECMI (Fundação Getúlio Vargas School of Communication, Media and Information) who studies social media platforms, also makes a “terrible” diagnosis. For the specialist, the situation can be harmful even for the social network.

He cites the monitor of the debate on voter fraud and digital distrust as one of the FGV projects to be impacted by the new policy. The tool, for internal use, is built on Twitter’s public API.

It made it possible to follow what the public was saying about topics such as interventionism, election security and printed voting. It was possible to identify, for example, whether the speech of a politician generated an increase in discussions on one of these topics.

“Today, life is increasingly digital and that means that the problems are also within digital environments. How do we think about solutions to fake news, misinformation, hate speech? This depends on research, on people who delve into the data to think about solutions and eventually demand actions from the platforms”, he says.

Piaia claims that the ease of access to data, something that is more complicated on competing platforms, caused a distortion that was beneficial to Twitter: it was more studied and, therefore, appeared as an influential and important network in social life, even with penetration smaller than rivals.

Another negative point is the use in education. Extracting information through the API can be a way to teach data analysis and programming. “It was used a lot in the classroom”, says the researcher.

The idea of ​​​​the change, according to Elon Musk, came to stop automated posts. One of the main negative marks of the social network is the avalanche of content produced by coordinated robots – in disinformation campaigns, for example.

“The free API is being exploited by fraudsters and opinion seekers. There is no verification process or cost, so it’s easy to create 100,000 robots to do bad things. A $100 monthly access should clean things up pretty well,” he wrote Musk.

A side effect, however, is a possible end of the line for innocent robots, since it is not known which ones could have the privilege of staying on the platform for free.

Enter this cake the funniest ones, like what automatically share images of SpongeBob cartoonor those who publish useful information on the platform.

One such case is Project 7c0. It monitors profiles of political actors on the social network and alerts, through tweets, when it detects that a post has been deleted.

“Currently, I believe it’s the end of the line, also because US$ 100 is a very high cost for a project with just one person”, says Lucas Lago, creator of the system.

Lago’s initiative relies on the API in three steps: to collect posts, check for removals, and publish deleted ones. The limitation of sharing 1,500 tweets per month could even be circumvented, explains the programmer, by placing a ceiling on daily posts.

“The problem is that we don’t have any information about the other two uses I make of the API. I can’t even find out if paying the US$ 100 would be possible to keep the project running”, says Lago.

Sought, Twitter did not respond to the report for clarification on the policy change.

The API change adds to a series of controversial changes implemented since Elon Musk took over the company and was not the first to target robots.

In December, he had already made a change to the social network’s policy to try to undermine the sharing of third-party location information. At the time, the pivot of the story was a profile that was banned for automatically sharing the whereabouts of the billionaire’s jet based on public data.

The social network’s week was busy, with users reporting difficulties posting tweets on Wednesday. The problem happened on the same day that the Twitter Blue subscription package for BRL 42 per month was announced in Brazil. The service cuts ads in half, guarantees a blue verified seal and allows you to publish texts of up to 4,000 characters.