Jack Dorsey, the co-founder of the Twitter that killed Vine in 2017, is now backing its revival as diVine, a six-second video platform with a strict ban on AI-generated content, Guardian Project AI detection tools built into the upload flow, and a founding proposition that human creativity is not a feature but the entire product.
The app is live. It launched to testers last November under Dorsey's nonprofit vehicle "and Other Stuff," went public in May 2026, and is currently operating with an archive of around 500,000 videos from the original Vine alongside new submissions from human creators. The lead operator is Evan Henshaw-Plath, known online as Rabble, an early Twitter employee who has been building toward a permanent home for Vine's content library for several years. Henshaw-Plath's pitch is that the original Vine represented a specific kind of internet creativity, constrained, fast, looping, and human, that the algorithmic feed era and the generative AI era have both eroded in different ways. diVine is an attempt to reassemble that format under a set of rules that prevent the erosion from happening again.
The enforcement mechanism matters because any platform can write a policy banning AI content and few can implement it credibly. diVine requires new uploads to either be recorded directly within the app or verified as human-made through a third-party verification tool. Guardian Project, the privacy and security software organisation, provides the AI detection layer that flags suspected synthetic content before it can be posted. That is a more concrete infrastructure investment than a community guideline, but it is not foolproof. AI video detection accuracy is improving but still produces false positives and false negatives at rates that would frustrate both legitimate human creators and determined bad actors. The honest version of diVine's content authenticity pitch is that it raises the cost and friction of posting AI-generated content significantly, not that it eliminates the possibility entirely. At a platform of this scale, that distinction is manageable. At the scale diVine needs to reach to compete meaningfully with TikTok or Reels, it becomes the central operational and technical challenge of the business.
The competitive positioning against AI-generated content is a real market signal rather than pure nostalgia marketing. Research from Pew and others has consistently shown that users across age groups are developing scepticism toward AI-generated media, particularly in entertainment and social contexts where authenticity and relatability are the primary value drivers. Short-form video is specifically the format where that scepticism is sharpest, because the appeal of a six-second loop has always been the sense that a real person did something funny or interesting or strange in front of a camera and shared it without the production apparatus of traditional media. Sora, Runway, and Kling have made credible six-second AI-generated video cheap enough that any platform without an enforcement mechanism will be flooded with synthetic content within months of achieving meaningful scale. diVine's founders understand that the credibility of the human-only label is what they are selling, and that the moment the label is convincingly disputed, the core value proposition collapses.
The Dorsey angle is more complicated than the press coverage suggests. Dorsey left Twitter in 2021 and has since positioned himself as a critic of the centralised platform model he helped build, investing in decentralised social infrastructure through Bluesky and related projects. Backing diVine through a nonprofit vehicle rather than a commercial entity is consistent with that positioning but also raises questions about the business model. diVine's guidelines mention integration with Dorsey's Block ecosystem for potential monetisation, which points toward creator payments and tipping infrastructure rather than advertising as the revenue mechanism. That approach is philosophically coherent but has not yet been demonstrated to generate the kind of revenue that sustains a content moderation team, a detection technology partnership, a server infrastructure, and a competitive creator acquisition budget simultaneously. The original Vine could not solve its creator monetisation problem even with 100 million monthly active users and Twitter's balance sheet behind it. diVine is starting with a fraction of that audience and no advertising engine.
The broader pattern diVine fits into is consumer startup positioning around authenticity as a response to AI content saturation. It is not the only company making this bet. BeReal built its brief cultural moment on a similar premise, that the curated inauthenticity of Instagram was a problem worth solving with a product constraint. Locket Widget, Poparazzi, and a wave of ephemeral and friend-first social apps have all experimented with format and audience constraints as substitutes for the algorithmic feeds that dominate the incumbents. None of them have scaled into durable platforms, and the common failure mode is that the constraint that creates early differentiation becomes a growth ceiling once the initial enthusiast cohort is exhausted and mainstream creators demand the flexibility and monetisation tools they already have on TikTok and Reels. diVine's six-second limit and its AI ban are both constraints in this sense. They create identity and early loyalty. They also limit the creator universe and the content surface area that drives discovery and retention.
The question for founders watching this launch is not whether diVine will beat TikTok. It will not. The question is whether the anti-AI positioning is a fundable thesis for a new category of consumer platforms, and whether that category has enough addressable audience to support a standalone business. The evidence from search trends, platform surveys, and the early creator response to diVine's launch suggests that the audience for human-only content is real and motivated. Whether it is large enough, loyal enough, and willing to pay enough to sustain a platform through the content moderation costs that honest enforcement of a human-only policy requires, is the calculation that will determine whether diVine becomes a durable product or a well-intentioned experiment that demonstrates the demand without capturing the economics.
Also read: IBM's MAMMAL Is a Quiet Demonstration That Biomedical AI Is Moving Beyond Single-Purpose Models • Panthalassa Wants to Build AI Data Centers in the Ocean and the Power Crunch Makes That Sound Less Crazy Than It Should • AI Models Are Giving Biosecurity Experts Operationally Useful Bioweapons Guidance and the Refusal Systems Are Not Stopping It
