How the biggest tech companies are building AI: language models, assistants and agents, vision, voice, and the systems that train them. The artificial intelligence and machine learning patent applications we have broken down from the companies we track, newest first.
Week of November 12, 2026 · 96 filings covered · refreshes every week
The 96 filings Patentlyze covered this week sort into four clusters: making AI faster and smaller for on-device use, generating and editing images and video, improving speech and language tools, and training AI models more efficiently.
Google has filed a patent application for an ad with a chat button that asks you to upload photos of your stuff, then recommends something to buy based on what it sees. The filing's main example: pictures of your apartment, and a moving truck sized to fit your furniture.
Circle a building in a photo, and this Google patent application describes an AI that turns it into a podcast with a synthetic host and guest, timed to fit your drive.
Disney has described a game world that watches the real world and changes to match it: a parade starts in a physical park, and a virtual plaza fills with floats and confetti at the same moment.
Nvidia's patent application describes an AI that turns a typed sentence into a full 3D CT, MRI or PET scan, then files that fake scan into a dataset used to train other medical AI. To make it understand everyday wording, another AI first rewrites real radiology reports in many styles.
Instead of retraining a speech AI for every new language, Google describes leaving it untouched and bending the incoming sound until a Lithuanian 'ne' looks like an English 'no.'
Every time you tell a chatbot 'that's wrong,' you hand its maker a free lesson. This Google patent describes a way to collect those lessons from phones without uploading the conversations themselves.
A new Nvidia patent application tackles a flaw in AI-animated faces: when a character talks fast, the mouth blurs sounds together. Its fix is to stretch the audio clues the AI sees, then keep only the middle.
AI image tools make you wait because the model builds a picture through many small passes. A new Qualcomm patent application describes a model that finishes in as few as two to four, by learning from several teacher models at once.
Say the wrong name in a podcast and today you have to re-record. Nvidia's filing describes retyping the word and letting software drop in a matching synthetic voice for just that spot.
Speech-to-text software stumbles on words it has rarely heard, like a drug brand, a company name or a small-town street. Amazon's filing describes handing the system a list of those words and nudging it to pick them.
Most translation tools try to translate everything they hear. This Nvidia patent application describes a model that listens to a messy conversation, finds the one person you care about, and translates only them.
Google's patent describes a way to fix a group photo by borrowing a better face, or a whole head, from another frame in the same burst. The twist is that it decides which kind of swap will look more natural.
Type 'Victorian wallpaper with lemons' and Adobe's patent filing describes software that wraps your text box in a matching pattern, shaped to fit it.
Qualcomm's patent decides which video frames an AI should bother with by comparing two sentences: one describing the whole clip, one describing a single frame. If they say about the same thing, the frame gets skipped.
Adobe's filing describes an AI that turns the letters you type into artwork, such as a bonsai tree shaped like an A, and keeps them as real text. Edit the word later and the style redraws itself.
Apple's patent application teaches face recognition to spot people behind masks by training on masks that were never actually worn.
2,637 filings, broken out by what the AI is built to do. Every type is its own tracked page.
Over the last 90 days, Google leads with 249 AI / ML filings, followed by Microsoft (180), Nvidia (154), Adobe (118) and IBM (107). That is out of 1,221 AI / ML filings from all the companies we track in the same period.
188 AI / ML patent applications from the companies we track have published in November 2026 so far, after 472 in October. Patentlyze has covered 2,637 AI / ML filings since May 2026.
Google Patents a Way to Teach Speech AI New Languages Without Retraining It, a Google filing published November 12, 2026. The latest filings are listed above, newest first.
No. A published application is not a product announcement. Companies file to protect options, and most patented ideas never ship in the form described. What filings do show is which companies are putting research effort into AI / ML right now.
