Alleged AI-Generated Photo Alteration Leads to Inappropriate Modifications in Speaker's Conference Picture

An AI image expansion tool used by a conference organizer unintentionally altered Elizabeth Laraki’s profile picture for a marketing ad, making her blouse appear unbuttoned and showing a fabricated hint of undergarments. The AI generated the lower part of the image when expanding it for vertical formatting. The conference organizers quickly apologized and removed the altered content.

Risk classification

• Primary risk domain: 1 Discrimination & Toxicity
• Primary risk subdomain: 1.2 Exposure to toxic content

The AI tool generated inappropriate, sexualized content by altering the subject's blouse to appear unbuttoned with visible undergarments, exposing the subject and viewers to toxic and inappropriate content.

Additional risk subdomains

• 1.1 Unfair discrimination and misrepresentation: The AI exhibited systemic gender bias by automatically sexualizing a professional woman's portrait, reflecting unfair representation and ingrained misogyny in its training.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The sexualized alteration was generated by an AI image expansion tool post-deployment, representing an unexpected and unintentional outcome of the user's goal to make the image taller.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The incident involves the generation of synthetic, altered media (deepfake-like image manipulation of a real person's likeness) which carries transparency and disclosure risks.

AI system and alleged parties

• AI system: AI expand image tool
• AI purpose: Image Generation; Image Retouching
• Behaviour type: Tool
• Alleged developer: Unknown developer
• Alleged deployer: Unknown conference employee
• Alleged harmed parties: Elizabeth Laraki

Harm severity

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

Psychological

Reported: The report explicitly describes psychological distress and shock experienced by the affected individual.

Directly caused: One person, [PERSON_001], experienced shock, distress, and felt 'horrified' upon seeing her professional photo altered to show a made-up bra.

Indirectly caused: N/A

Inferred additional harm: N/A

People affected

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

Potential causes

Management

• Lack of AI Usage Policies: Organizers lacked policies regarding acceptable use of AI editing tools.
• Insufficient Quality Oversight: Management failed to oversee social media content generation workflows.

Technology

• Biased Generative Outpainting: AI tool sexualized the female subject during image expansion.
• Inaccurate Feature Generation: AI removed pockets and necklace medallion while inventing clothing details.

Data Inputs

• Lack of Original Vertical Photo: Social media manager lacked the original vertical image for the ad.
• Biased Training Data: Training data ingrained a sexualized male-gaze bias for women's apparel.

Human Factors

• Uncritical Trust in AI Output: Social media manager used the AI tool without verifying the output.
• Lack of AI Literacy: Staff did not understand generative AI's tendency to hallucinate and bias.

Process and Methods

• Inadequate Image Review Process: No formal review of promotional materials before posting them online.
• Use of Generative Fill for Headshots: Using AI outpainting on professional portraits without strict guidelines.

Regulatory Environment

• Absence of AI Safety Standards: No industry standards governing generative AI bias in commercial tools.

Information quality

• Classification confidence: High
• Reason for confidence: The reports are highly consistent, featuring direct quotes and first-hand accounts from the affected individual, [PERSON_001], explaining exactly how the AI tool was used and how the organizers responded.
• Ambiguities identified: The specific brand or name of the AI expander tool used by the social media manager is not explicitly named.
• Alternative interpretations: None, the facts of the AI's generation of the altered image and the subsequent correction are clear and undisputed.

An AI image expansion tool inappropriately altered a professional portrait, sparking discussions on algorithmic gender bias. The incident is a localized, unintentional software error with negligible national security implications.

• Overall national security impact: Negligible
• Response level: Minor
• Scope: Single nation
• Primary target: No clear primary
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. The incident reflects ongoing societal and ethical concerns with AI bias rather than an active national security crisis.
• Autonomy: Human-controlled. The AI functioned as a tool used by a human operator, who retained control over publishing the final output.
• Novelty: Established threat. Biased or inappropriate outputs from generative AI models are well-documented and represent an established technical limitation.

Impact by dimension

• Physical security: Negligible. The incident involves an AI image expansion tool and has no impact on physical systems, infrastructure, or kinetic capabilities.
• Information security: Negligible. No classified systems were breached, and there is no evidence of state-sponsored disinformation or intelligence compromise.
• Sovereignty: Negligible. The incident does not affect state authority, electoral processes, or core government operations.
• Economic security: Negligible. No strategic technology theft or financial system manipulation occurred; the economic impact is negligible.
• Societal stability: Negligible. While the incident highlights gender bias in AI, it is an isolated event affecting a single individual and does not threaten large-scale social cohesion.