Topline
OpenAI launched its newest video generation model Tuesday with the release of “Sora 2,” introducing the ability to add audio and users’ likeness into generated videos—one of the latest artificial intelligence innovations made this year as titans like OpenAI, xAI, Google and Anthropic compete for AI dominance.
Open AI CEO Sam Altman speaks during a talk session with SoftBank Group CEO Masayoshi Son at an event titled “Transforming Business through AI” in Tokyo on Feb. 3, 2025. (Photo by Tomohiro Ohsumi/Getty Images)
Getty Images
Key Facts
Sora 2 improves on its predecessor by allowing users to incorporate synchronized dialogue and sound effects to the video clips generated by the model.
OpenAI simultaneously launched an app for Sora, adding a social media component to the video generator that lets users remix each other’s generations and add the likeness of themselves or their friends into videos with a feature known as “cameos.”
Users can insert their likeness into videos after making a video-and-audio recording in the Sora app to verify their identity.
The Sora app was immediately made available for download in the U.S. and Canada, with ChatGPT Pro users receiving access to an “experimental, higher quality” Sora 2 model on Sora’s own website.
Get Forbes Breaking News Text Alerts: We’re launching text message alerts so you’ll always know the biggest stories shaping the day’s headlines. Text “Alerts” to (201) 335-0739 or sign up here.
What Are The Latest Chatbots From Some Of The Biggest Ai Firms?
Anthropic launched Claude Sonnet 4.5, its latest AI model, in September, featuring improvements to coding and business-related requests. OpenAI’s newest language model, GPT-5, launched in August featuring quicker responses and a lower likelihood of hallucinations (when a model accidentally gives incorrect or misleading results). Gemini 2.5 Pro, Google’s newest model, was launched in June and incorporated better-organized responses that include the use of headers, lists and tables to make complicated user requests easier to understand. xAI’s latest Grok model, Grok 4, was released in July with improved search integration and problem-solving skills. Meta’s Llama 4 model released in April, boasting cheaper costs for inputs and outputs and larger context windows, which are key to models maintaining lengthy conversations and handling large requests from users.
What Are The Latest Ai Video Generators?
Meta’s AI video generation offerings were expanded in September, when the company launched a preview of Vibes, a social app for sharing AI videos with friends and followers. Meta’s video generation software is known as Meta Movie Gen, which uses text inputs to create custom videos. Like Sora 2, users of Meta Movie Gen can use their likeness in generated videos by uploading a photo of themselves (Sora 2 requires video-and-audio recordings from users to incorporate their likeness into videos). Vibes and Meta Movie Gen are free. xAi’s video generating model is known as Grok Imagine, which can generate short, six-second videos with sounds from text inputs. The model is free, though premium users can get access to longer videos, higher resolution and higher usage limits. Google’s video generation model, VEO 3, was launched in May and features text-to-video and image-to-video generations.
The Latest Ai Valuations
OpenAI leads the pack in terms of estimated value. The Sam Altman-led company was valued at $300 billion following a March funding round and is reportedly eyeing a secondary share sale that would value the startup at $500 billion—a figure that would make OpenAI the world’s most valuable privately held company. Elon Musk’s xAI raised $10 billion in debt and equity in September, valuing the company around $150 billion. However, the company quickly positioned itself to improve on the figure just a week after the raise, reportedly raising a separate $10 billion from investors that would bring its estimated worth to $200 billion. Anthropic raised $13 billion in a September funding round, valuing the startup at $183 billion.
Who Are Some Of The Billionaires Most Active In The Ai Boom?
The AI boom has made longtime tech billionaires richer and introduced new billionaire tech moguls. Jensen Huang, the chief executive of chip designer and AI investor Nvidia, has boomed in net worth in the last two years, reaching an estimated $162 billion and making him the world’s seventh-wealthiest person as of September. Anthropic’s sibling co-founders Dario Amodei and Daniela Amodei became billionaires this year, with both reaching $3.7 billion estimated net worths. Altman became a billionaire last year and has a $2.2 billion estimated net worth attributable to investments in Reddit, Stripe and nuclear fusion company Helion (he holds no equity in OpenAI). Musk was a billionaire well before establishing xAI and is the world’s wealthiest person, boasting a $490 billion estimated net worth in September. Musk and Altman, whose startups are direct competitors, have publicly feuded several times, scrutinizing each other’s company policies, products and business approaches. Musk has characterized Grok as an anti-“woke” version of OpenAI’s ChatGPT and said OpenAI is “a closed source, maximum-profit company effectively controlled by Microsoft.” Altman said on a podcast in 2023, “Most of that is not true, and I think Elon knows that.” Musk made an unsolicited $94.7 billion offer for OpenAI early this year. Altman said the company was “not for sale” and that Musk’s offer was meant to “slow us down.”
Big Number
$109 billion. That is how much money was privately invested into AI last year, according to a report from the Stanford Institute for Human-Centered Artificial Intelligence.
Key Background
The Stanford report also concluded more businesses are using AI (78% of organizations reported AI use in 2024, up from 55% in 2023). The report noted improvements to language models has made AI cheaper to use and more accessible, with hardware operating costs dropping 30% annually and energy efficiency improving by 40% each year. Altman told Axel Springer Global Reporters Network in September that he would be surprised if AI superintelligence was not achieved by 2030. AI superintelligence refers to hypothetical AI systems smarter than humans.