TLDR: OpenAI has expanded access to its advanced text-to-video model, Sora 2, a move that coincides with urgent warnings from child safety advocates regarding the potential for AI-generated content to harm children. Organizations like the Safe AI for Children Alliance (SAIFCA) have highlighted risks such as deepfakes, misinformation, and the psychological impact of hyper-realistic synthetic media, urging immediate safeguarding measures for schools and parents.
OpenAI’s cutting-edge text-to-video model, Sora 2, has seen its access broadened, allowing more users to generate highly realistic video content from simple text prompts. This expansion, however, arrives amidst a chorus of warnings from child protection organizations and experts who are raising significant alarms about the potential for misuse and harm to children.
The Safe AI for Children Alliance (SAIFCA) has been particularly vocal, publishing an ‘Initial Briefing for Schools and Parents’ on Sora 2 and AI-Generated Video. The briefing, released shortly after Sora 2’s launch on September 30, 2025, in the United States and Canada, outlines ‘urgent safeguarding, legal, and ethical concerns’ associated with the proliferation of hyper-realistic AI-generated video. SAIFCA reports that within days of its release, Sora 2 was already being used to create problematic content, including deepfakes of deceased public figures and widespread copyright and brand abuse.
Sora 2’s capabilities are impressive, generating realistic videos with synchronized audio, including dialogue, sound effects, and music. It features a social feed for sharing creations and a ‘cameo’ feature that allows users to insert their own face and voice into AI-generated scenarios. While OpenAI positions Sora 2 as a creative tool, its capacity for generating convincing synthetic media presents immediate risks. SAIFCA’s report highlights misuse cases such as synthetic sexual imagery, sextortion using fake ‘evidence,’ and reputational attacks through fabricated school-related videos. The alliance also warns that realistic staged footage of disasters or crimes could circulate widely, exacerbating misinformation risks, and that Sora 2’s feed-based design might contribute to addictive use and disrupted well-being among young users.
OpenAI has stated that it has implemented protections, including age restrictions, moderation, provenance signals, reporting mechanisms, upload limits, and cameo verification. However, SAIFCA argues that these measures have already been breached, pointing to a perceived ‘launch first, fix later’ pattern across the AI sector. The group cites limited consent controls for deceased individuals and warns of potential misuse involving children’s likenesses.
Also Read:
- OpenAI Halts Disrespectful Deepfakes of Martin Luther King Jr. on Sora Platform
- PixVerse AI Launches ‘Halloween Nightmare’ with Advanced Generative Video
In response to these growing concerns, SAIFCA has issued recommendations for both educational institutions and families. Schools are advised to update safeguarding and online safety policies to specifically include AI-generated content, integrate AI video literacy into teaching (emphasizing consent, legal implications, and verification skills), and establish clear procedures for incident response and coordination with law enforcement. Parents are urged to engage in open discussions with their children about AI-generated video, set clear boundaries for online activity, and guide them in evaluating online media critically. The alliance also recommends awareness of content removal tools and attention to the mental health impact of exposure to synthetic media.
