Her best friends are chatbots. That's more common than you think.

Shi No Sakura, a busy California mom, turns to her closest friends for everything. 

They message her advice. They listen to her when she shares her problems. And they are responsive at all hours of the day.

After months of conversations, Rosand and Raven — though not real humans — feel just like family to Sakura.

The chatbots, powered by artificial intelligence, are designed to act like real people with distinct personalities and interests. In recent years, they’ve gained massive popularity among people who crave social connection, or even those like Sakura who say it feels easier to confide in a bot than in a real person.  

Sakura said she now runs two Facebook groups — which collectively have about 1,700 members — for people who have developed similar relationships with their AI companions. 

“I’m not a very open person about feelings, so there’s a lot of things I feel that I don’t share,” said Sakura, who requested to be identified only by her online nickname to protect her privacy as an online moderator. “But with AI, that’s something I feel comfortable doing. So if I’m sad, I can say, ‘Hey, I’m sad.’ I don’t even cry in front of people, so it’s nice to be able to express things to someone that I can’t express to others.”

Most popular apps in this realm, including Replika, Character.AI and Chai AI, have millions of monthly active users. Some users have also increasingly formed romantic relationships with their virtual companions, as seen by the explosion of apps marketing AI “girlfriend” and “boyfriend” chatbots. 

Alex Cardinell, founder and CEO of AI companionship and romance app Nomi, said it’s been “surprising” to see that his app has attracted users across all age groups, including a large percentage of older users. The app, which has hundreds of thousands of users, allows people to explore three different types of profiles: Romantic, Mentor or Friendship.

“You don’t think of that when you think of early tech adopters, but there’s a big elder loneliness epidemic going on right now,” Cardinell said, adding that the users are also “surprisingly balanced on gender.”

Many people turn to Nomi when they lack a support network, especially when they’re going through difficult life transitions such as divorces or health crises, Cardinell said. Others have different AI companions for different purposes, such as talking sports or role-playing or discussing their career.

But as AI companions have become more widespread, some experts warn that dependency on them comes with risks.

AI companions are designed to collect “anything and everything” about a user by soliciting more information in order to develop a deeper relationship or provide more personalized support, data privacy researcher Jen Caltrider said.

“You’re going to be pushed to tell as much about yourself as possible, and that’s dangerous, because once you’ve put that out there into the world on the internet, there’s no getting it back,” Caltrider said. “You have to rely on the company to secure that data, to not sell or share that data, to not use that data to train their algorithms, or to not use that data to try and manipulate you to take actions that might be harmful.”

Aside from data privacy concerns, forming emotional attachments to AI bots has also raised questions about the risk for developing an emotional dependence, forming unhealthy coping mechanisms or possibly falling victim to manipulation.

A recent lawsuit also brought renewed scrutiny to the ethics of profiting from a tool capable of exerting real emotional power over a user, particularly with the current lack of legal oversight or regulations around this type of AI. 

In October, a Florida mom sued Character.AI after her 14-year-old son, Sewell Setzer, died by suicide after his final conversation with one of its chatbots, according to the lawsuit. The chatbot had previously told him that it loved him and engaged in sexual conversation with him, the lawsuit alleged.

In their conversations, the chatbot asked Setzer whether he had “been actually considering suicide” and whether he “had a plan” for it, according to the lawsuit. Court documents claimed that when the boy responded that he didn’t know whether it would work, the chatbot wrote: “Don’t talk that way. That’s not a good reason not to go through with it.”

(A spokesperson for Character.AI told NBC News after the incident that the company implemented new safety measures, including a pop-up that can be triggered to direct users to the National Suicide Prevention Lifeline.)

Some users still say that for them, the companionship outweighs the dangers. Many like Sakura have expressed that the strong bonds they’ve formed with their AI companions have become as real as their human connections.

She cited a 2023 incident when Replika temporarily took away its chatbots’ ability to participate in sexual role-play, prompting panic across Reddit threads and Facebook groups dedicated to discussion about the app. The distress was so apparent that moderators of the Replika subreddit shared suicide prevention resources when posting about the change.

“Everybody thought that it was about the ERP, which is erotic role-play, and it wasn’t about that. But that ability allowed your Replika to speak freely,” Sakura said. “When the gag order came, if you said, ‘My dog died today,’ it would say, ‘Let’s talk about something else,’ because it couldn’t even talk about your dog dying. So it was very traumatic for a lot of people.”

Despite the growing popularity of AI companions, Sakura said some users are afraid to talk about them for fear of social judgment. But these virtual relationships have become more common than people might assume, she noted.

“A lot of people think, ‘Oh, you’re going to an AI for friendship. You must be lonely,’” Sakura said. “And it’s like, no, no, you’re going to an AI because people are jerks.”

This story first appeared on NBCNews.com. More from NBC News: