The findings suggest a major uptick in suspected bot activity online since Lula’s October win in Brazil, which like the United States has been besieged by election fraud claims.
According to a report by the Israeli firm Cyabra shared with The Technology 202, researchers estimated that the percentage of Twitter accounts believed to be “inauthentic” discussing the election grew steadily after October, multiplying fivefold to upward of 20 percent by January.
The firm scanned nearly 10,000 profiles mentioning the Brazilian election since September across Twitter and Facebook, finding similar but less pronounced trends on the latter.
Cyabra spokesperson Ilan Fisher said they use machine-learning to scan for characteristics in thousands of accounts and posts to see if they are “displaying nonhuman behavior.” They then assign scores and use statistical analysis to determine if accounts are likely “inauthentic.”
According to the report, the Twitter posts shared by the suspected fake accounts were overwhelmingly anti-Lula: roughly 1 in 5 referred to the Brazilian president as a “thief” or mentioned election rigging claims.
Researchers found that the accounts’ most used hashtag overall was #ForaLuladrao, which roughly translates to “Out with Lula the thief,” and that the growing prevalence of fake accounts after October coincided with a significant boost in the exposure for that phrase online.
In total, the firm estimated that the accounts reached tens of millions of users across Twitter and Facebook, with the bulk of that happening on Twitter.
An email to Twitter’s communications department, which was gutted after Musk’s takeover, brought no response. Musk did not return a request for comment. Facebook declined to comment.
The findings raise questions about the effectiveness of Musk’s push to eliminate bots, which he made a top priority long before taking over the reins at Twitter.
Early on in his bid to acquire the company, Musk pledged to “defeat the spam bots or die trying!”
Musk later made it a central argument in his bid to back out of the deal, claiming the company had misled him in downplaying how much of the site was made up of bots.
Just how much of Twitter’s user base is made up of bots is hotly contested.
Cyabra said it was commissioned by Musk in May to look into the issue and has previously estimated that 13.7 percent of Twitter’s user base is made up of “inauthentic” profiles. (Fisher said the firm is not currently working with Twitter, without elaborating.)
Other researchers have claimed the percentage is lower and in the single-digits. And while many of the models used to calculate the rates are sophisticated, they are still estimates.
The apparent uptick in bot activity also puts a spotlight on Twitter’s enforcement outside of the United States, where it has slashed its moderation workforce even more dramatically.
As my colleague Lizza Dwoskin reported last week, Musk “fired the company’s entire staff in Brazil except for a few salespeople.” Those fired included “eight people, based in São Paulo who moderated content on the platform to catch posts that broke its rules against incitement to violence and misinformation.”
Both the staffing cuts and the suspected increase in bot activity largely preceded the violent attacks on Brazil’s government buildings earlier this month led by supporters of former president Jair Bolsonaro, some of whom allege the election was stolen.
Social networks have long faced pressure to crack down on spam and bots, particularly given the history of governments using them for disinformation campaigns.
But there are also still lingering questions about how effective those tactics have been in shaping minds, if at all.
According to a report by our friends from The Cybersecurity 202, a recent study found that Russia’s infamous influence operations on Twitter in 2016 “reached relatively few users” and “had no measurable impact in changing minds or influencing voter behavior.”
Our top tabs
Trump’s presidential campaign asks Facebook to reinstate him
A lawyer for former president Donald Trump’s 2024 presidential campaign said in a letter to Mark Zuckerberg, the CEO of Facebook’s parent Meta, and other company executives that Facebook’s two-year ban of Trump’s account has “dramatically distorted and inhibited the public discourse,” John Wagner reports. Scott Gast, the lawyer, called for a meeting “to discuss President Trump’s prompt reinstatement to the platform.”
“Meta suspended Trump indefinitely on Jan. 7, 2021,” John writes. “The company’s Oversight Board, an independent group of human rights experts, academics and lawyers that issues binding rulings on some of Meta’s content moderation decisions, later upheld the suspension but criticized the company for not establishing criteria for suspending a user indefinitely. The company then shortened the suspension to two years and said that when that period was over, it would assess whether the public safety risk had subsided enough to restore his account.”
Meta spokesman Andy Stone said last week that the company will announce its decision “in the coming weeks in line with the process we laid out.”
OSHA cites Amazon over workplace hazards in three warehouses
The citations by the Department of Labor’s Occupational Safety and Health Administration (OSHA) pertained to hazards at three Amazon warehouses in New York, Illinois and Florida, CNBC’s Annie Palmer reports. OSHA said the workers were “exposed to ergonomic hazards” and were at a high risk of being affected by injuries and other disorders related to their work. Amazon got a separate citation for allegedly exposing some workers at one of the warehouses to falling boxes.
“Each of these inspections found work processes that were designed for speed but not safety, and they resulted in serious worker injuries,” Assistant Secretary for Occupational Safety and Health Doug Parker said in a statement. “While Amazon has developed impressive systems to make sure its customers’ orders are shipped efficiently and quickly, the company has failed to show the same level of commitment to protecting the safety and well-being of its workers.”
Amazon has long been scrutinized for the treatment of its workers and conditions of its warehouses. Spokesperson Kelly Nantel told CNBC in a statement that the company disagrees with OSHA’s findings and plans to appeal. “We’ve cooperated fully, and the government’s allegations don’t reflect the reality of safety at our sites,” Nantel said.
(Amazon founder Jeff Bezos owns The Washington Post.)
Microsoft’s layoffs add to industry cuts
Microsoft, which announced layoffs of 10,000 employees, is the latest tech giant to announce reductions in its workforce amid economic uncertainty and higher interest rates, Jacob Bogage, Hamza Shaban and Taylor Telford report. Microsoft’s reliance on cloud computing and ubiquitous software for revenue make it especially plugged in to business trends.
“Microsoft laying off is a canary in the coal mine for the broader economy,” said Dan Ives, managing director at Wedbush Securities. “And I think it shows that enterprises are starting to slow spending, with a mild recession likely on the doorstep.”
Major technology companies have announced plans to cut at least 57,000 jobs in the coming months. Amazon also kicked off a round of cuts Wednesday, trimming its workforce by more than 18,000 people.
- Michael Kennedy is joining Intuit as its chief corporate affairs officer. Kennedy previously worked as senior vice president for global government relations and public policy at VMware and as chief of staff to former senator Orrin G. Hatch (R-Utah).
- NTIA Director Alan Davidson speaks at the U.S. Conference of Mayors 91st Winter Meeting today.