‘100 Daily Video Calls’: Models Seeking Roles as the Face of AI Scams

Other postings indicate up to 150 potential calls each day. “Filters may be utilized, but make sure the image is realistic. Live-action videos are acceptable; wigs are not,” states another advertisement. In exchange, individuals would supposedly receive one full day and four half days off every month. Additionally, one listing specifies work hours from 10 pm to 10 am in Cambodia, with a preference for candidates to have a “Western accent.” A job ad for models mentions: “The company will hold onto your passport for visa and work permit management.” Taking passports is a primary method used by scam operators to control individuals.

While a few men are applying for the AI model positions, the majority of applications reviewed by WIRED came from young women, primarily in their early twenties. Candidates are requested to submit a brief video introduction, written details about their experiences and expectations, along with photographs; some are even asked to provide their marital status and “vaccination” status.

“For over three years, I have collaborated with Chinese firms on various projects including stock market, cryptocurrency, and romance narratives,” states one individual in a recruitment video. Another claims: “Based on my experience, I excel at managing clients, persuading them to invest using my own methods and discussing the advantages of gold trading.”

The video applications lack full names or contact information, preventing WIRED from reaching out to the applicants.

Modeling applicants have sought salaries as high as $7,000 per month, as per Humanity Research Consultancy. They also make particular requests regarding their working conditions, many of which may not be available to those trafficked into the scam operations. One woman asked for her own room and the ability to “go outside.” Another requested they could “return home on their day off” and have a “personal washing machine.”

Though some models may have more freedoms than victims of human trafficking, says Ling Li, cofounder of the nonprofit EOS collective assisting victims of the scam industry, they may still endure harsh treatment from superiors. “One European victim told us he observed some Italian models in his compound, but he could not ascertain whether they were there willingly as he witnessed them being beaten,” she recounts. “Additionally, there have been reports of sexual harassment.”

WIRED provided Telegram with a list of two dozen job and recruitment channels advertising AI models, among other roles, in recent months. The company appeared to leave these channels active; however, a spokesperson asserted that their policies prohibit any scam-related activities.

“Content that promotes or facilitates scams is explicitly banned by Telegram’s terms of service and is removed whenever it is identified,” a spokesperson for Telegram stated. “In situations like these, there are legitimate reasons one might provide their likeness, so such content has to be evaluated on a case-by-case basis.”

The vast majority of model-job advertisements and applications on Telegram do not overtly mention scam-related work, yet they contain numerous red flags indicative of scamming activities, according to Ngo. “Why [do you] need AI models? That’s the first question,” Ngo emphasizes. Other warning signs include locations known for scamming in Cambodia, claims of exorbitant salaries for the area, and frequent requirements for proficiency in Chinese, Ngo adds.