Every Christmas growing up in Minnesota, Jimmy Darts’ parents gave him $200 in cash: $100 for himself and $100 for a stranger. Now, with over 12 million followers on TikTok and several million more on other platforms, philanthropy is his full-time job.

Darts, whose real surname is Kellogg, is one of the biggest creators of “kindness content,” a subset of social media videos devoted to helping strangers in need, often with cash amassed through GoFundMe and other crowdfunding methods. A growing number of creators like Kellogg give away thousands of dollars – sometimes even more – on camera as they also encourage their large followings to donate.

“The internet is a pretty crazy, pretty nasty place, but there’s still good things happening on there,” Kellogg told The Associated Press.

Not everyone likes these videos, though, with some viewers deeming them, at their best, performative, and at their worst, exploitative.

Critics argue that recording a stranger, often unknowingly, and sharing a video of them online to gain social media clout is problematic. Beyond clout, content creators can make money off the views they get on individual videos. When views reach the millions, as they often do for Kellogg and his peers, they make enough to work full-time as content creators.

Comedian Brad Podray, a content creator formerly known online as “Scumbag Dad,” creates parodies designed to highlight the faults he finds with this content — and its proponents — as one of the most vocal critics of “kindness content.”

“A lot of young people have a very utilitarian mindset. They think of things only in measurable value: ‘It doesn’t matter what he did, he helped a million people’,” Podray said.

Recording practices prompt questions of ethics

From the recording devices and methods down to the selection of subjects, “kindness content” — like everything on social media — exists on a spectrum.

Some creators approach strangers and ask them for advice or for a favor, and if they bite, they receive a prize. Others choose to reward strangers they see doing a good deed. Kellogg performs a “kindness challenge,” asking a stranger for something and returning it in kind.

Many of these strangers are unaware they’re being filmed. Some creators employ hidden cameras and aim to record subjects in a discreet manner. Kellogg said he wants to be as “secret about it as possible,” but asks for consent to share the video after the interaction. Kellogg said most agree because they look “like a superhero” after his challenge.

Another charitable content creator, Josh Liljenquist, said he uses a GoPro camera and tries to make recording “extremely noticeable,” adding, “Consent’s the biggest thing.”

Regardless of the recording method, some see the process as predatory.

“These guys always find someone with cancer or always find someone who can’t pay their bills because they’re stalking through underserved and poor areas and they’re just sort of waiting,” Podray said. “Looking through the parking lot like, ‘He looks pathetic enough’.”

Karen Hoekstra, the marketing and communications manager for the Johnson Center for Philanthropy, studies TikTok-based influencer philanthropy and says the videos, at times, take advantage of their subjects.

“The model of the man on the street walking up and approaching a stranger and handing them money is — we’ve all heard this phrase, terrible as it is — it just strikes me as poverty porn,” Hoekstra said. “It’s exploitation.”

Calls of exploitation often come when creators feature the same people across multiple videos, especially when they appear to be homeless or have a drug addiction. Liljenquist features some people frequently and maintains that his recurring subjects are like his “best friends.”

One user commented on an Oct. 5 video that recent content feels like Liljenquist is “playing case worker for views,” as he posted several videos of a woman who followers suspect is struggling with a drug addiction. He records himself bringing her food, giving her a ride in his Tesla, and asking her questions that often get one-word responses.