Background
MITII, Inc. sued OpenAI in October 2025, alleging that OpenAI’s Sora text-to-video application infringes four of MITII’s patents. The patents-in-suit relate to systems originally developed to help children with autism — technology that converts electronic text messages into speech using animated characters’ voices and transmits the speech synchronized with stored moving images. MITII argued that OpenAI’s Sora, which generates video content from text prompts using AI models, infringes these patents.
The case was originally filed in the Central District of California and subsequently transferred to the Northern District of California. OpenAI moved to dismiss, arguing that MITII’s patents do not cover OpenAI’s AI-powered video generation technology.
The Court’s Holding
Judge Noël Wise granted OpenAI’s motion to dismiss without leave to amend, effectively ending the case. The court found that MITII’s patent claims, as applied to OpenAI’s Sora platform, could not survive a motion to dismiss, and that amendment would be futile — a significant determination given that courts typically grant pro se or small plaintiffs at least one opportunity to replead.
The dismissal with prejudice means MITII cannot refile these claims against OpenAI based on the same patents, marking a definitive win for OpenAI in this particular patent challenge.
Key Takeaways
- AI video generation technology like OpenAI’s Sora faces ongoing patent litigation from holders of older multimedia patents — but courts are drawing clear lines between legacy text-to-animation systems and modern AI-driven video synthesis.
- Dismissal without leave to amend signals the court found a fundamental disconnect between MITII’s patent claims and OpenAI’s technology that no amount of repleading could cure.
- This is one of several patent suits OpenAI faces as its product suite expands, but the dismissal with prejudice provides a clean resolution on these particular patents.
Why It Matters
As AI companies like OpenAI expand into video generation, image creation, and other creative tools, they face a growing wave of patent litigation from holders of earlier-generation multimedia technology patents. This case illustrates the challenges plaintiffs face in asserting legacy patents against fundamentally different AI architectures: while both the patented technology and Sora involve converting text into visual media, the underlying technical approaches are worlds apart. The decisive dismissal with prejudice sends a signal to other potential patent plaintiffs that merely covering a similar functional space does not make older multimedia patents applicable to modern AI systems.