Nina Tomasieski logs on to Twitter before the sun rises. Seated at her dining room table with a nearby TV constantly tuned to Fox News, the 70-year-old grandmother spends up to 14 hours a day tweeting the praises of President Trump and his political allies, particularly those on the ballot this fall, and deriding their opponents.

She’s part of a dedicated band of Trump supporters who tweet and retweet Keep America Great messages thousands of times a day.

“Time to walk away Dems and vote RED in the primaries,” she declared in one of her voluminous tweets, adding, “Say NO to socialism & hate.”

While her goal is simply to advance the agenda of a president she adores, she and her friends have been swept up in an expanded effort by Twitter and other social media companies to crack down on nefarious tactics used to meddle in the 2016 election.

And without meaning to, the tweeters have demonstrated the difficulty such crackdowns face — particularly when it comes to telling a political die-hard from a surreptitious computer robot.

Last week, Facebook said it had removed 32 fake accounts apparently created to manipulate U.S. politics — efforts that may be linked to Russia.

Twitter and other sites also have targeted automated or robot-like accounts known as bots, which authorities say were used to cloak efforts by foreign governments and political bad actors in the 2016 elections.

But the screening has repeatedly and erroneously flagged Tomasieski and users like her.

Their accounts have been suspended or frozen for “suspicious” behavior — apparently because of the frequency and relentlessness of their messages. When they started tweeting support for a conservative lawmaker in the GOP primary for Illinois governor this spring, news stories warned that right-wing “propaganda bots” were trying to influence the election.

“Almost all of us are considered a bot,” says Tomasieski, who lives in Tennessee but is tweeting for GOP candidates across the U.S.

Cynthia Smith has been locked out of her account and “shadow banned,” meaning tweets aren’t as visible to others, because of suspected “automated behavior.”

“I’m a gal in Southern California,” Smith said. “I am no bot.”

The actions have drawn criticism from conservatives, who have accused Twitter, Facebook and other companies of having a liberal bias and censorship. It also raises a question: Can the companies outsmart the ever-evolving tactics of U.S. adversaries if they can’t be sure who’s a robot and who’s Nina?

“It’s going to take a really long time, I think years, before Twitter and Facebook and other platforms are able to deal with a lot of these issues,” said Timothy Carone, who teaches technology at Notre Dame’s Mendoza College of Business.

The core problem is that people are coming up with new ways to use the platforms faster than the companies can manage them, he said.

Twitter did not respond to a request for comment. But the company has said it identified and challenged close to 10 million suspected bot or spam accounts in May, up from 3.2 million last September. It’s also trying to weed out “trolls,” or accounts that harass other users, pick fights or tweet material that’s considered inflammatory.

Twitter acknowledges that there will be some “false positives.”

“Our goal is to learn fast and make our processes and tools smarter,” Twitter executives said in a blog post earlier this year.

Tomasieski and her conservative friends use so-called Twitter “rooms” — which operate using the group messaging function — to amplify their voices.

She participates in about 10 rooms, each with 50 members who are invited in once they hit a certain number of followers. That number varies, but “newbies” might have around 3,000, Tomasieski says. Some have far more.

Everyone in the room tweets their own material and also retweets everyone else’s. So a tweet that Tomasieski sends may be seen by her roughly 51,000 followers, but then be retweeted by dozens more people, each of whom may have 50,000 or more followers.

She says she’s learned some tricks to avoid trouble with Twitter. She’s careful not to exceed limits of roughly 100 tweets or retweets an hour. She doesn’t use profanity and she tries to mix up her subjects to appear more human and less bot-like.

During a recent afternoon, Tomasieski retweeted messages criticizing immigrants in the U.S. illegally, Democratic socialists and the media. One noted an Associated Press story about an increase in the number of Muslims running for public office — news the user described as “alarming.”

Tomasieski says she loves to write. But most important is helping “my guy.”

“There is as much enthusiasm today as there was when Trump was elected. It’s very quiet, but it’s there. My job is to get them to the polls,” she said. “That’s rewarding. I go to bed feeling like I have accomplished something.”
…

Nina Tomasieski logs on to Twitter before the sun rises. Seated at her dining room table with a nearby TV constantly tuned to Fox News, the 70-year-old grandmother spends up to 14 hours a day tweeting the praises of President Trump and his political allies, particularly those on the ballot this fall, and deriding their opponents.

She’s part of a dedicated band of Trump supporters who tweet and retweet Keep America Great messages thousands of times a day.

“Time to walk away Dems and vote RED in the primaries,” she declared in one of her voluminous tweets, adding, “Say NO to socialism & hate.”

While her goal is simply to advance the agenda of a president she adores, she and her friends have been swept up in an expanded effort by Twitter and other social media companies to crack down on nefarious tactics used to meddle in the 2016 election.

And without meaning to, the tweeters have demonstrated the difficulty such crackdowns face — particularly when it comes to telling a political die-hard from a surreptitious computer robot.

Last week, Facebook said it had removed 32 fake accounts apparently created to manipulate U.S. politics — efforts that may be linked to Russia.

Twitter and other sites also have targeted automated or robot-like accounts known as bots, which authorities say were used to cloak efforts by foreign governments and political bad actors in the 2016 elections.

But the screening has repeatedly and erroneously flagged Tomasieski and users like her.

Their accounts have been suspended or frozen for “suspicious” behavior — apparently because of the frequency and relentlessness of their messages. When they started tweeting support for a conservative lawmaker in the GOP primary for Illinois governor this spring, news stories warned that right-wing “propaganda bots” were trying to influence the election.

“Almost all of us are considered a bot,” says Tomasieski, who lives in Tennessee but is tweeting for GOP candidates across the U.S.

Cynthia Smith has been locked out of her account and “shadow banned,” meaning tweets aren’t as visible to others, because of suspected “automated behavior.”

“I’m a gal in Southern California,” Smith said. “I am no bot.”

The actions have drawn criticism from conservatives, who have accused Twitter, Facebook and other companies of having a liberal bias and censorship. It also raises a question: Can the companies outsmart the ever-evolving tactics of U.S. adversaries if they can’t be sure who’s a robot and who’s Nina?

“It’s going to take a really long time, I think years, before Twitter and Facebook and other platforms are able to deal with a lot of these issues,” said Timothy Carone, who teaches technology at Notre Dame’s Mendoza College of Business.

The core problem is that people are coming up with new ways to use the platforms faster than the companies can manage them, he said.

Twitter did not respond to a request for comment. But the company has said it identified and challenged close to 10 million suspected bot or spam accounts in May, up from 3.2 million last September. It’s also trying to weed out “trolls,” or accounts that harass other users, pick fights or tweet material that’s considered inflammatory.

Twitter acknowledges that there will be some “false positives.”

“Our goal is to learn fast and make our processes and tools smarter,” Twitter executives said in a blog post earlier this year.

Tomasieski and her conservative friends use so-called Twitter “rooms” — which operate using the group messaging function — to amplify their voices.

She participates in about 10 rooms, each with 50 members who are invited in once they hit a certain number of followers. That number varies, but “newbies” might have around 3,000, Tomasieski says. Some have far more.

Everyone in the room tweets their own material and also retweets everyone else’s. So a tweet that Tomasieski sends may be seen by her roughly 51,000 followers, but then be retweeted by dozens more people, each of whom may have 50,000 or more followers.

She says she’s learned some tricks to avoid trouble with Twitter. She’s careful not to exceed limits of roughly 100 tweets or retweets an hour. She doesn’t use profanity and she tries to mix up her subjects to appear more human and less bot-like.

During a recent afternoon, Tomasieski retweeted messages criticizing immigrants in the U.S. illegally, Democratic socialists and the media. One noted an Associated Press story about an increase in the number of Muslims running for public office — news the user described as “alarming.”

Tomasieski says she loves to write. But most important is helping “my guy.”

“There is as much enthusiasm today as there was when Trump was elected. It’s very quiet, but it’s there. My job is to get them to the polls,” she said. “That’s rewarding. I go to bed feeling like I have accomplished something.”
…