Yazılar

OpenAI to Give Content Owners Control Over Sora AI Videos, Plans Revenue Sharing Model

/in /tarafından

OpenAI is rolling out new tools to give content owners greater control over how their intellectual property is used in Sora, its recently launched AI video-generation app, and plans to introduce a revenue-sharing system for creators who opt in.

In a blog post on Friday, CEO Sam Altman said OpenAI will soon provide “more granular control over the generation of characters” within Sora, enabling rights holders such as film and television studios to decide how their characters can appear—or to block them entirely.

The move comes amid intensifying scrutiny of AI-generated content and growing concern across Hollywood and the creative industries about copyright infringement and the unauthorized replication of proprietary characters and likenesses.

Sora, launched this week as a standalone app in the United States and Canada, allows users to generate and share AI-created videos up to 10 seconds long. Its social-media-style interface quickly gained traction, with users producing clips based on both original and copyrighted material.

Altman acknowledged that the app’s rapid popularity—and the sheer volume of video creation—has outpaced expectations, creating a need for clear rules and compensation mechanisms. “We’ll experiment with different approaches,” he wrote, adding that the revenue-sharing model would evolve through “trial and error” as OpenAI tests various systems within Sora before applying them to its broader suite of AI tools.

At least one major studio, Disney, has already opted out of allowing its characters to appear in Sora-generated videos, sources familiar with the matter told Reuters. Other studios are reportedly reviewing whether to participate under OpenAI’s forthcoming licensing framework.

The company’s initiative could mark a turning point in the relationship between AI firms and content owners, shifting from conflict to collaboration—if a viable monetization model can be found.

Backed by Microsoft, OpenAI’s expansion into multimodal AI via Sora places it in direct competition with Meta’s Vibes and Google’s text-to-video tools, as major tech firms race to define the future of synthetic media creation.

Still, the effort to give rights holders control over how their creations are used—and to share revenue from those uses—reflects a broader recognition that AI’s creative power must coexist with creator compensation and consent.

Jeff Bezos Envisions Gigawatt-Scale Data Centres in Space Within Two Decades

/in /tarafından

Amazon founder Jeff Bezos says it’s only a matter of time before humanity builds massive data centres in orbit, powered by continuous solar energy and free from Earth’s environmental constraints. Speaking at Italian Tech Week in Turin, Bezos predicted that gigawatt-scale orbital data hubs could become a reality within the next 10 to 20 years, eventually surpassing their terrestrial counterparts in efficiency and cost.

“These giant training clusters—those will be better built in space,” Bezos said during a conversation with Ferrari and Stellantis Chairman John Elkann. “We have solar power there 24/7—no clouds, no rain, no weather. We will be able to beat the cost of terrestrial data centres in space in the next couple of decades.”

The idea of space-based computing infrastructure is gaining traction among tech firms as AI-driven demand for electricity, cooling, and server capacity skyrockets on Earth. Conventional data centres are now among the world’s largest industrial consumers of energy and water, fueling the search for sustainable alternatives.

Bezos described orbital data centres as a natural next step in the broader trend of using space to improve life on Earth, noting that satellites already manage weather forecasting, communications, and navigation. “The next step is data centres, then other kinds of manufacturing,” he said.

However, the vision faces formidable obstacles: high launch costs, maintenance difficulties, and the risk of mission failures in space. Frequent upgrades—a routine part of Earth-based data infrastructure—would be far more complicated in orbit.

Beyond technology, Bezos framed the discussion within a broader narrative about AI and societal transformation. Drawing parallels between today’s artificial intelligence boom and the dot-com bubble of the early 2000s, he urged optimism despite speculative excess.

“We should be extremely optimistic that the societal and beneficial consequences of AI—like we had with the internet 25 years ago—are for real and there to stay,” he said. “It’s important to separate potential bubbles from the actual underlying reality.”

Bezos emphasized that AI’s impact will be “broadly diffused” across industries and societies, suggesting that the technology’s true promise lies not in isolated breakthroughs but in its global, everyday applications.

His comments add weight to the emerging idea that space infrastructure could become the next great frontier of the digital economy, where data, energy, and AI converge far above Earth’s atmosphere.

Brazilian Police Bust Deepfake Scam Using Gisele Bündchen’s Image in Instagram Ads

/in /tarafından

Brazilian authorities have dismantled a nationwide fraud network that used deepfake videos of supermodel Gisele Bündchen and other celebrities in Instagram ads to trick victims into buying fake products, marking one of the country’s first major crackdowns on AI-powered online scams.

Police arrested four suspects this week and froze assets across five states, after investigators traced more than 20 million reais ($3.9 million) in suspicious transactions uncovered by Brazil’s anti–money laundering agency COAF.

The investigation began in August 2024, when a victim reported being deceived by an Instagram ad showing an AI-generated video of Bündchen promoting a nonexistent skincare product. Another fraudulent campaign featured the supermodel supposedly offering free suitcases, with users asked to pay only for shipping—items that never arrived.

According to Eibert Moreira Neto, head of the cybercrime unit in Rio Grande do Sul, the group created a “series of scams” using deepfakes of multiple celebrities and fake betting platforms. Investigators believe the criminals operated at mass scale, collecting many small payments—usually under 100 reais ($19)—from victims who rarely reported the losses.

“That created a perverse situation,” explained investigator Isadora Galian. “The criminals enjoyed a kind of statistical immunity—they knew most people would not complain, so they operated without fear.”

Meta, owner of Instagram, said its policies ban ads that deceptively use public figures and that such content is removed “when detected.” The company added that it uses AI-based detection systems, trained review teams, and reporting tools to fight celebrity-impersonation scams.

A spokesperson for Bündchen’s team urged consumers to verify suspicious offers, avoid ads promising unrealistic discounts or giveaways, and report fraudulent content to authorities or official brand channels.

The case has broader implications for Brazil’s fight against digital deception. In June 2024, the Supreme Court ruled that social media platforms can be held liable for criminal ads if they fail to remove them swiftly—even without a court order.

The Rio Grande do Sul operation underscores the growing criminal use of deepfake technology, which allows scammers to replicate celebrity likenesses with stunning realism. What once required Hollywood budgets can now be done with cheap AI tools and a few clicks—a reality that’s forcing regulators, platforms, and the public to confront a new era of synthetic fraud.