For the arrest related to the creation, possession, and transportation of child sexual abuse material (CSAM), which includes utilizing web-based artificial intelligence software to generate pornographic images of minors, a child psychiatrist named David Tatum, 41, was apprehended on Wednesday.
Following a jury’s conviction in May, David Tatum was sentenced to 40 years in prison, with an additional 30 years of supervised release. Moreover, he was mandated to pay $99,000 in compensation for his crimes.
US Attorney Dena J. Tatum, who specializes in child psychiatry, emphasized Tatum’s awareness of the severe and enduring impact of sexual abuse on children’s well-being. Despite this knowledge, Tatum engaged in the reprehensible practice of creating illicit images and videos from the private recordings of his patients.
Dena J. Tatum affirmed her commitment to prosecuting individuals who exploit technology to harm children, condemning Tatum’s misuse of artificial intelligence to victimize minors.
The indictment against Tatum does not delve into specifics regarding the AI software employed. However, additional court documents revealed that Tatum not only viewed AI-generated inappropriate images of children on a deep fake platform but also possessed, produced, and distributed sexually explicit material involving minors.
Trial evidence presented by the prosecution included covert recordings of minors, including a cousin, undressing and showering, as well as engaging in sexual activities.
Furthermore, trial evidence highlighted Tatum’s use of web-based artificial intelligence tools to morph clothed images of adolescents into explicit content. This underscores the alarming trend of utilizing AI technology to transform images of clothed minors into child pornography.
A recent CNN report detailed a case in South Korea where an individual received a two-and-a-half-year prison sentence for producing physical images of children.
The increasing concern surrounding the use of AI for creating CSAM has garnered attention from legislators, civil society groups, and AI service providers.
OpenAI CEO Sam Altman disclosed that GPT-4 exhibits an 82% decreased likelihood compared to GPT-3.5 in responding to requests for prohibited content. Altman also mentioned employing a rigorous evaluation process involving both human and automated assessments to prevent misuse of AI technologies.
Altman further highlighted the utilization of Thorn’s Safer support by OpenAI to detect, block, and report instances of CSAM.
However, efforts to identify CSAM post-creation due to network monitoring obligations could potentially compromise online security. Organizations like Thorn have advocated for legislation mandating online content monitoring to combat CSAM, emphasizing the need for such services.
