Deepfake Images of Australian Teacher Hannah Grundy and 25 Others Created and Circulated Online by Former Friend

Australian teacher Hannah Grundy discovered her face had been superimposed onto pornographic images using AI deepfake technology, which were shared online alongside personal details. A former trusted friend, Andrew Thomas Hayler, was found responsible, targeting 26 women, including Grundy and close friends, over two years. Hayler pleaded guilty to multiple charges and was sentenced to nine years in prison.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.3 Fraud, scams, and targeted manipulation

The perpetrator used AI to create humiliating and sexual deepfake imagery of specific victims to harass and degrade them.

Additional risk subdomains

• 1.2 Exposure to toxic content: The incident involved the creation and online distribution of highly toxic, violent, and sexually explicit content depicting rape fantasies.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The incident was caused by a human actor intentionally using deployed AI deepfake generation tools to harass and abuse victims.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The incident involves AI-generated content such as deepfakes, which under the EU AI Act are classified as limited risk, requiring transparency obligations.

AI system and alleged parties

• AI system: unspecified
• AI purpose: Deepfake Image Generation; Deepfake Video Generation
• Behaviour type: Tool
• Alleged developer: Unknown deepfake technology developers
• Alleged deployer: Andrew Thomas Hayler
• Alleged harmed parties: Hannah Grundy, Former friends and acquaintances of Andrew Thomas Hayler, Colleagues of Hannah Grundy

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Minor
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Minor, indirect Minor
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Minor, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Minor, indirect Negligible
• Psychological: direct Minor, indirect Minor
• Epistemic: direct Minor, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: Yes, Hannah Grundy and Kris Ventura spent $20,000 on legal fees and forensic experts to build the case.

Indirectly caused: The victims spent approximately $20,000 on legal representation and forensic computing experts to gather evidence and compel police action.

Inferred additional harm: Other victims may have incurred unquantified financial costs related to therapy, security systems, or lost productivity due to distress.

Malicious content

Reported: Yes, deepfake pornography and violent rape fantasies were uploaded online.

Directly caused: The perpetrator created and uploaded hundreds of sexually explicit deepfake images and videos of 26 women to a pornography website with over 300,000 members.

Indirectly caused: The content continued to circulate and resurface on other websites even after the original forum was decommissioned.

Inferred additional harm: The synthetic explicit content will likely remain accessible on various corners of the internet indefinitely, causing ongoing exposure.

Civil rights

Reported: Yes, violations of privacy, bodily autonomy, and safety.

Directly caused: The perpetrator violated the victims' rights to privacy, personal safety, and dignity by generating and distributing non-consensual sexual imagery of them.

Inferred additional harm: The ongoing online presence of the deepfakes represents a permanent violation of the victims' right to control their own likeness and privacy.

Privacy

Reported: Yes, the perpetrator stole photos from private social media accounts and leaked personal details.

Directly caused: The perpetrator violated the privacy of 26 women by taking photos from their private social media accounts and publishing them alongside their full names, occupations, and locations.

Psychological

Reported: Yes, the reports explicitly describe severe psychological trauma, anxiety, panic attacks, and feeling unsafe.

Directly caused: The 26 victims suffered profound psychological distress, anxiety, and trauma. Hannah Grundy experienced panic attacks, withdrew from social life, and felt terrified to leave her house.

Indirectly caused: Partners and friends of the victims, such as Kris Ventura, experienced significant stress and anxiety from discovering the website and investigating the perpetrator.

Inferred additional harm: It is highly likely that all 26 victims suffered long-term trauma, anxiety, and paranoia regarding their personal safety and online presence.

Epistemic

Reported: Yes, the creation of highly realistic fake images and videos that falsely depicted the victims engaging in sexual acts.

Directly caused: The AI-generated deepfakes fabricated realistic but entirely false depictions of the victims in degrading and violent pornographic scenarios.

People affected

• Occurrences reported: 1
• People reportedly harmed: 26
• People reportedly exposed: 26

Potential causes

Management

• Unregulated Hosting Platforms: Forum operators failed to manage and moderate illegal user-uploaded content.
• Inadequate Social Media Privacy: Social platforms failed to prevent friends from downloading private photos.

Technology

• Accessible Deepfake AI Tools: AI technology allowed easy superimposing of faces onto graphic images.
• Realistic AI Graphic Editing: AI added realistic details like bruising to make images highly disturbing.

Data Inputs

• Social Media Image Scraping: Private social media photos were downloaded and used as inputs for AI models.

Human Factors

• Perpetrator Psychological Issues: Addictions to violent porn and substances fueled the malicious behavior.
• Misperception of Online Harm: Perpetrator dismissed the severe real-world impact as just online harm.
• Abuse of Interpersonal Trust: A close friend exploited his access to victims' private profiles.

Process and Methods

• Inadequate Initial Police Response: Police initially dismissed evidence, delaying arrest and prolonging trauma.
• Lack of Platform Content Moderation: The host website allowed graphic non-consensual deepfakes and rape polls.
• Manual Forensic Requirements: Victims had to spend $20,000 on private investigators to prove the crime.

Regulatory Environment

• Gaps in Cyber Abuse Legislation: Initial lack of specific laws for AI-generated non-consensual pornography.
• Slow Law Enforcement Action: Lack of established protocols for handling digital image-based abuse cases.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts of the perpetrator's actions, the legal proceedings, the sentencing, and the specific impacts on the victims. Multiple reputable news sources corroborate the facts of the case.
• Ambiguities identified: The specific AI software or models used by the perpetrator are not named, referred to only generally as 'AI technology' or 'deepfake technology'.
• Alternative interpretations: None. The facts of the criminal case and the role of deepfake technology are clear and undisputed.

An Australian man was sentenced to nine years in prison for creating and distributing non-consensual deepfake pornography of 26 women. While the incident represents a severe, harmful violation of individual privacy and human rights, it remains a localized criminal matter with negligible national security implications.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: Australia
• Alleged perpetrator: Andrew Hayler

Threat characteristics

• Imminence: Long-term. The threat is a resolved domestic criminal case, representing ongoing societal and legal challenges rather than an active national security crisis.
• Autonomy: Human-controlled. The AI system was used as a tool by the perpetrator, who manually gathered source photos and directed the creation of the deepfakes.
• Novelty: Established threat. Non-consensual deepfake pornography is a known and established threat, though this case represents a significant domestic prosecution and sentencing precedent.

Impact by dimension

• Physical security: Negligible. No physical security, critical infrastructure, or kinetic systems were targeted or affected in this incident.
• Information security: Negligible. While deepfakes were generated, they targeted private individuals for harassment rather than constituting state-sponsored information warfare or intelligence operations.
• Sovereignty: Negligible. No government functions, electoral systems, sovereign boundaries, or state decision-making processes were impacted.
• Economic security: Negligible. The incident represents a localized criminal case and does not threaten financial systems, strategic technologies, or economic security.
• Societal stability: Minor. Represented severe violations of individual privacy and human rights for 26 victims, but remained localized and manageable within standard domestic criminal justice systems.