The rapid democratization of generative artificial intelligence has introduced a volatile new element into the educational environment, transforming traditional schoolyard bullying into a sophisticated form of digital character assassination. In recent months, a surge of incidents involving the creation and dissemination of nonconsensual, AI-generated sexual imagery of teachers by their students has surfaced across the United States and the United Kingdom. These deepfakes, often indistinguishable from reality at a cursory glance, are being used as weapons of harassment, leaving educators traumatized and school administrators struggling to adapt existing disciplinary frameworks to a landscape where the tools of defamation are available to any teenager with a smartphone.
The Case of Luis DeSantiago: A Chronology of Digital Contagion
The experience of Luis DeSantiago, a substitute teacher in Los Banos, California, serves as a quintessential example of how quickly these digital attacks can escalate. On March 12, while working a second job at a coffee shop, DeSantiago was alerted by a colleague to an AI-generated photo circulating on TikTok. The image was a sophisticated manipulation of a mirror selfie DeSantiago had taken years prior during college. In the altered version, his professional attire was replaced with lingerie, and his posture was modified to be suggestively sexual.
The timeline of the incident reveals a cascade of institutional failures and the viral nature of AI content:
- March 12: Discovery of the image. DeSantiago identifies a local middle school student as the primary source based on TikTok profile data and social media connections.
- March 13: DeSantiago reports the incident to the assistant principal of the middle school where he was scheduled to work. He discovers a second TikTok account impersonating him, using his name and the deepfake as a profile picture.
- Mid-March: Formal investigations begin. A student admits to creating the image and receives an in-school suspension. However, the image continues to proliferate on Snapchat and TikTok.
- Late March – April: The harassment crosses physical boundaries. While subbing at a local high school, the same image is "AirDropped" to DeSantiago during a class session, indicating that the content had successfully jumped from one student body to another.
- May: Despite administrative offers of counseling, DeSantiago makes the decision to withdraw from the district, citing a lack of physical and psychological safety.
This chronology highlights a critical issue in modern school administration: once an AI-generated image enters the digital ecosystem of a student body, traditional disciplinary measures like individual suspensions are often insufficient to halt its spread.
The Psychological Toll and the "Uncanny Valley" of Harassment
The impact of these incidents extends beyond professional reputation; it strikes at the core of the educator-student relationship. An anonymous middle school teacher in her mid-twenties shared a similar account involving a TikTok profile that used AI to simulate her voice and likeness. The deepfakes depicted her skydiving and performing backflips, but also included audio of her describing explicit sexual acts.
The teacher noted that the "smoothness" of the AI-generated skin and the unnatural movements of the mouth—often referred to as the "uncanny valley" effect—did not prevent younger students from believing the content was authentic. The psychological distress was compounded when the student responsible for the videos transitioned to verbal, in-person harassment, calling her derogatory names and refusing to follow instructions.
In the United Kingdom, a youth worker named Alice experienced a similar betrayal of trust. A student she had mentored—discussing sensitive topics like safe sex and personal safety—was suspected of creating a deepfake video of her undressing. The incident fundamentally altered her approach to her vocation. While she previously prided herself on having a strong rapport with students, she now describes the atmosphere as "chilled," necessitating a more guarded and vigilant professional distance.
Data and the Proliferation of Nonconsensual Deepfakes
While comprehensive national data on AI-driven teacher harassment is still being compiled, broader trends in digital safety indicate a burgeoning crisis. According to reports from cybersecurity firms specializing in deepfake detection, such as Sensity AI, over 90% of deepfake videos found online are nonconsensual pornography. While historically targeting celebrities, the technology has now been "democratized" to the point where private individuals, including public servants like teachers, are increasingly targeted.
A 2023 survey by the National Education Association (NEA) found that nearly 20% of educators reported being the victim of some form of cyberbullying or digital harassment. The introduction of generative AI tools like Midjourney, Stable Diffusion, and various "undressing" apps has lowered the barrier to entry for creating malicious content, making it a primary tool for student-led harassment.
Legal Frameworks and the "Take It Down Act"
The legal landscape regarding AI-generated nonconsensual imagery is currently a patchwork of state and federal initiatives. In the United States, the "Take It Down Act" (S.3696) represents a significant federal attempt to criminalize the publication of nonconsensual "intimate visual depictions," whether they are authentic or computer-generated. This legislation aims to provide victims with a clearer path to seeking justice and forcing platforms to remove content.
However, legal experts like Donna Ballman, a retired employment advocacy lawyer, point out that existing federal antidiscrimination laws already place a burden on employers. Under Title IX and various workplace safety regulations, schools are obligated to provide a workplace free of sexual harassment. If a student creates a sexualized image of a teacher, it constitutes a "clear case" of sexual harassment. The challenge for schools lies in the "no blueprint" reality: administrators are often unsure how to balance student privacy and disciplinary records with the victim’s right to know how the situation is being handled.
Institutional Responses and Platform Accountability
Social media platforms remain at the center of this crisis. While companies like Snap and TikTok have community guidelines prohibiting harassment and nonconsensual sexual content, the enforcement of these rules is often reactive rather than proactive. In the case of Luke Red, an Arizona-based sixth-grade teacher, TikTok initially failed to remove accounts that were circulating an AI-generated photo of him kissing another male teacher. Red was forced to report the accounts for being under the minimum age of 13 rather than for the content itself to see any action taken.
School districts have responded with varying degrees of severity. In Red’s district, the student responsible received a three-day suspension—the maximum allowed for a first-time sexual harassment offense under the student handbook. Many educators argue that these punishments are disproportionately light compared to the long-term professional and personal damage caused by the digital dissemination of sexualized deepfakes.
Broader Implications for the Teaching Profession
The rise of AI-driven harassment arrives at a time when the teaching profession is already facing a recruitment and retention crisis. The added layer of digital vulnerability is likely to exacerbate this trend.
1. Erosion of the Mentor-Student Bond
As seen in the case of Alice in the UK, the fear of being "deepfaked" encourages teachers to maintain a sterile, clinical distance from their students. This defensive posture undermines the mentorship and emotional support that are often vital to student success, particularly for at-risk youth.
2. The Persistence of the Digital Record
Unlike traditional rumors, which fade over time, AI-generated images can be archived, re-uploaded, and circulated indefinitely. This "permanent record" of harassment can follow a teacher throughout their career, potentially affecting future employment opportunities if a background check surfaces malicious, albeit fake, content.
3. The Digital Literacy Gap
Many students do not fully grasp the legal or ethical ramifications of their actions. Luke Red noted that his students seemed to operate in a "permissive culture" regarding AI, viewing the technology as a toy rather than a tool for potential defamation. This suggests an urgent need for updated digital citizenship curricula that specifically address the ethics of generative AI.
Conclusion: A Call for Systemic Change
The victimization of educators like Luis DeSantiago, Luke Red, and others underscores a systemic vulnerability in the education sector. Addressing this issue requires a three-pronged approach: stronger legislative protections that specifically address AI-generated harassment, more robust and proactive moderation from social media platforms, and a comprehensive overhaul of school disciplinary codes to reflect the severity of digital sexual harassment.
As AI technology continues to evolve, the window for creating effective safeguards is closing. Without decisive action, the classroom—once a space for growth and trust—risks becoming a primary theater for the most invasive forms of digital abuse, further driving talented educators away from a profession already under siege. The safety of the modern workplace now depends as much on digital firewalls and AI ethics as it does on physical security, a reality that schools can no longer afford to ignore.
About the Author
