Publié : 1 October 2025
Actualisé : 1 month ago
Fiabilité : ✓ Sources vérifiées
Je mets à jour cet article dès que de nouvelles informations sont disponibles.
OpenAI, the company behind ChatGPT, pushes AI boundaries with “Sora “—a new social network for AI-generated videos , and its “Sora 2” engine. Dubbed the “deepfake TikTok,” this platform offers an experience where all content is artificially generated from user images, raising crucial ethical and societal questions.
💰 Unveiling Sora: OpenAI’s Latest Innovation
Sora presents as a vertical-scrolling, short-video application, a popular format. Its unique feature: all content is 100% AI-generated by OpenAI, with no human influencers. This revolutionary approach promises total immersion in artificial creations.
At its core is “Sora 2,” OpenAI’s powerful video generation engine, converting text descriptions (“prompts”) into realistic animated scenes. This engine, a competitor to Google’s Veo 2, is essential to Sora, making AI content creation accessible.
💰 How Sora Works: Deepfakes as User Content
After avatar creation, users stage their digital double in AI-generated videos, up to 10 seconds. Early reviews describe “stunning” deepfakes, though generation can take several minutes.
The “Remix” feature is key: it allows other users to generate videos using your avatar. This is consensual deepfake, with OpenAI ensuring strict control over permissions.
“Transforming a formerly feared technology, deepfakes, into a tool for social creation is a bold gamble. Success will depend on the trust users place in the safeguards implemented.” — IActualité.
Currently in closed beta, the application is available on iPhone in the US and Canada. This phase allows OpenAI to refine the experience and security before wider deployment, addressing inherent challenges.
💰 Navigating the Ethical Minefield: OpenAI’s Safeguards
OpenAI has integrated security measures. Users control and can delete remixed videos of their avatar. Public figures require a “cameo” and explicit permission for image use.
Strict restrictions prohibit generating “X-rated or extreme” content, aiming to prevent abuse and ensure a safe environment. The effectiveness of these filters against virality remains questionable.
Deepfakes have long been perceived as a serious threat for misinformation. By integrating them at the core of a social network, OpenAI attempts to rehabilitate this technology, transforming a societal problem into a fun feature.
💰 Broader Implications and The Future of Digital Identity
A fundamental question emerges: who wants to engage with a social network where all content is AI-generated? The appeal for authenticity and real human interactions could be a major barrier to Sora’s adoption.
Sora’s launch is part of an era where AI redefines our relationship with image and digital identity. It blurs the lines between real and synthetic, offering new creative possibilities alongside fresh challenges for trust.
Sora is a bold experiment by OpenAI to popularize a controversial technology. Its success will depend on user adoption and the effectiveness of safeguards against misuse. The future of digital identity and social interaction might hinge on this platform.
❔ Frequently Asked Questions
How does the ‘strict consent’ mechanism for the Remix feature concretely protect users against unwanted uses of their deepfake avatar?
What are the specific societal implications of a platform where all content is 100% AI-generated and there are no human influencers, beyond the ethical challenges related to classic deepfakes?
Given the absence of human influencers and entirely AI-generated content, how does OpenAI envision Sora’s economic model to generate revenue, beyond technological innovation?
Since there are no human influencers on Sora, how does the platform anticipate the emergence of trends or the creation of a dynamic and engaging social culture?
🎥 Explanatory Video
Video automatically selected to enrich your reading
📋 Sommaire
📋 Sommaire
0 Comments