Confessore said it was challenging to find the right formula for "a really short form of storytelling," in a text message, but noted that the bot allowed readers to "drill down further" to get more details.
Phelps said bot usage has been "in six figures" with an audience that is younger and more global than Times readers.
"It's an effort to meet people on messaging platforms," he said. "We wanted to make it more personal, more interactive, to allow readers to feel more connected to the journalists themselves."
While bots offer no immediate monetization, it can helps bring more people to the newspaper's apps and website.
"This gets at the heart of relationship building," Phelps said.
The Washington Post meanwhile has its own bot using a robot icon, and separately will be using artificial intelligence technology to update dispatches on election night. The Post, owned by Amazon founder Jeff Bezos, developed the system called Heliograf that will help create hybrid human- and computer-generated stories.
Heliograf allows the Post "to create stories that are better than any automated system but more constantly updated than any human-written story could be," said Jeremy Gilbert, director of strategic initiatives at the newspaper.
BuzzFeed experimented with its "BuzzBot" on Messenger during the 2016 political conventions as part of an effort to better connect with readers and participants at the events. Amanda Hickman, who heads the BuzzFeed Open Lab, said bots offer "a one-to-one relationship which gives us the opportunity to let people fine-tune the product they want."
RELATED: Gorgeous Computer-Generated Flowers Bloom: Photos
Algorithmic systems which turn data into news stories have been used for several years, mainly for routine corporate results and minor league sports, but are now playing a role in election coverage.
The nonprofit news site ProPublica's election data bot, created with Google News Lab, updates every 15 minutes with election forecasts, campaign finance reports, Google Trends and other data. Another nonprofit called the PollyVote Project delivers similar dispatches based on poll results and other data.
"Every time we get new data we create an automated news item," said Andreas Graefe, a fellow at Columbia University's Tow Center for Digital Journalism who is leader of the project funded by the Tow Center and Germany's LMU Munich University. "We can publish articles seconds after we receive the data, and we can do it in an unlimited amount."
Graefe said several studies in Europe have shown that readers cannot tell the difference between a human- or computer-generated news article.
"When you ask people how readable a story is, they rate the human article better, but when you ask them how credible it is, the computer is better," he said. "We don't really know why."
RELATED: In Mock Elections, Who Wins the Student Vote?
Damian Radcliffe, a University of Oregon professor who follows automated journalism, said the trends over the current election season highlight gains made with technology.
"Bots and automation are increasingly becoming a part of how journalism is produced and content is being consumed," he said.
Radcliffe noted that automated journalism has moved beyond simple formulaic content to projects such as the Los Angeles Times "homicide report" covering every murder victim and its "quakebot" which delivers breaking news on earthquakes.
Radcliffe said many news organizations should look at these technologies even when resources are tight. For election coverage, "they provide and opportunity to publish information more quicky than humans can," he said.
While bots are not likely to replace reporters anytime soon, "they can free people up or allow parts of newsroom to do other things that could be valuable," Radcliffe said.