The organization first demonstrated Sora’s abilities in February 2024. In the intervening months, they’ve built a faster version and discovered how to responsibly release AI video generators.
OpenAI’s emphasis on security around Sora is standard for it. Generative AI Nowadays, however, it also shows the importance of precautions regarding AI that can be used to create convincing fake images, which can damage an organization’s reputation for example.
By December 10, an account had to be created on Soura. off Because of High demand.
What is Sora?
Sora is a creative AI diffusion model. Sora can create multiple characters, complex backgrounds, and realistic-looking movements in videos up to a minute long. It can also create multiple shots within a video, keeping the characters and visual style consistent and making Sora an effective storytelling tool.
Sora can be used to create videos with content, promote content or products on social media, or illustrate points in business presentations. While it shouldn’t replace the creative minds of professional video makers, Sora can be used to make some content more quickly and easily.
“Media and entertainment verticals will likely be early adopters of such models,” Gartner analyst and distinguished VP Arun Chandrasekaran Chandrasekaran told TechRepublic in an email in February. “Business functions like marketing and design within technology companies and enterprises can also be early adopters.”
The UK, Switzerland, and parts of Europe will not currently have access to Sora.
Currently, Sora is available with ChatGPT access in every region except the United Kingdom, Switzerland, and the European Economic Area. The Guardian pointed out that Sora still needs to comply with the EU’s GDPR and the Digital Services Act and the UK’s Online Safety Act. Open AI said In December it plans to expand access “in the coming months”.
How can I reach Sora?
By December, Chat GPT Plus and Pro Users can access Sora at sora.com.
All videos can be in 1080p resolution, up to 20 seconds long, and in widescreen, vertical, or square aspect ratios. The interface allows users to enter their own content, and a “storyboard” tool helps users organize their cues.
How does Sora work?
Sora is a diffusion model, meaning it gradually transforms a nonsensical image into a comprehensible image based on a prompt and uses a transformer architecture. The research done by OpenAI to create its DALL-E and GPT models – specifically the retrieval techniques from DALL-E – was the stepping stone to the creation of Sora.
See: Chief AI Officers May be key in APAC in 2025.
Sora’s videos don’t always look realistic.
Sora still has trouble telling right to left or following complex descriptions of events over time, such as clues about a certain camera movement. OpenAI said in February that videos created with Sora could be viewed through errors in cause and effect, such as someone biting into a cookie but not leaving a bite mark.
For example, interactions between characters can show blur (especially around limbs) or uncertainty in terms of numbers (eg, how many wolves are there at any given time in the video below?).
What are OpenAI’s security precautions around Sora?
With the right cues and tweaking, Sora’s videos can easily be mistaken for live action. OpenAI is aware of potential defamation or misinformation issues arising from this technology. The company said in December that it had safeguards in place to prevent “child sexual exploitation content and sexual deep faxes”. Usually people are “limited” to uploads.
If Sora is released to the public, OpenAI plans to watermark content created with Sora. C2PA Metadata Metadata can be viewed by selecting the image and choosing the File Info or Properties menu options. People who create AI-generated images can still remove metadata intentionally or accidentally.
OpenAI currently has nothing to stop users of its image generator. DALL-E 3by removing metadata.
“OpenAI’s decision to delay public access to Sora, despite the opportunity to release it sooner, is certainly commendable,” Nana Nwachukwu, an AI ethics and governance consultant at Saidot, said in an email to TechRepublic.
However, he said, it is too early to say how effective OpenAI’s mitigation strategy will be or if it will be rolled out in the EU.
“Technology as well as governance must be developed to monitor and manage these risks,” said Nwachukwu. “Without continuous monitoring and strong industry standards, the promise of innovation is overshadowed by the risk of misinformation and loss.”
“It already is [difficult] And increasingly AI-generated content will be impossible to detect by humans,” Chandrasekaran said in February. “VCs are investing in startups building deepfake detection tools, and they (deepfake detection) tools) can be part of an enterprise’s armor. However, in the future, identifying machine-generated content will require public-private partnerships, often at the time of creation.”
What are Sora’s competitors?
Sora’s photorealistic videos are quite different, but similar services exist. Perhaps the highest personality among them. Google Viewnow in private preview, and coming to Amazon Nova Rails.
The runway Provides enterprise-ready text-to-video AI generation. Flicki Social media can create limited videos with voice sync for narration. Generative AI can now reliably add or edit content to traditionally captured videos.
On February 8, Apple researchers revealed a paper about it. Key Framer’s proposed large language model that can create stylized, animated images.
Editor’s note: This article was originally posted in February and updated in December.