IBM Patent Locks AI Models Inside an Encrypted Vault From All Parties
What happens when two companies want to use each other's AI without trusting each other? IBM has a patent for exactly that…
27 filings · category
AI safety patents
Watermarks, guardrails, moderation, and model protection: the filings where Big Tech works out how to keep AI honest, and how to prove what's fake.
Recent filings
27 totalWatermarks, guardrails, moderation, and model protection: the filings where Big Tech works out how to keep AI honest, and how to prove what's fake.
Part of AI/ML · also in ai/ml: Language AIEnterprise AIAI visionAI assistant & agentAI training & infrastructureAI image & video
What we’re watching
Sony Patent Splits Personal Data Across Machines for Private AI Processing
Sony is patenting a way to run AI analysis on personal data without any single computer ever getting to see the whole…
Google Patents a Content Moderation System That Looks Up Live Context Before Acting
Most automated content moderation systems make decisions using only what they were trained on months ago. Google's new patent describes a system…
IBM Patents a System That Baits Its Own AI Chatbots Into Saying Harmful Things
IBM's new patent describes a system that deliberately tries to get an AI chatbot to say something harmful, and then uses every…
IBM Patents a System for Grading AI-Written Articles Against Human-Edited Ones
How do you know if an AI wrote something good enough? IBM is patenting a method that answers that question by measuring…
IBM Patents a Way to Protect AI Models by Feeding Them Controlled Noise
IBM wants to make AI models forget things they shouldn't remember, and the method involves deliberately corrupting the math happening inside the…
Google Patent Automatically Generates Ad Copy From Website Content
Google has patented a system that automatically reads a website, summarizes it, and uses that summary as a guardrail for writing ad…
Google Patent Turns Live Location Data Into Plain-English Travel Summaries
Instead of staring at a moving dot on a map, you'd get a sentence: 'Sarah is on the highway, about 20 minutes…
IBM Patents Software That Uses Cooperating AI Models to Spot Sensitive Data
Instead of trusting a single AI model to spot sensitive information, IBM's new patent describes a panel of AI models that check…
Google's New Patent Makes Private AI Training Far Less Expensive to Run
Training an AI model while keeping individual people's data genuinely private is expensive, and Google thinks it's found a way to make…
Google's New Patent Stops Its Smart Speakers From Talking to Themselves
If you've ever watched a Google Nest speaker respond to something it just said out loud, you've found the bug this patent…
Adobe Patents a System That Invents Fake Users to Train Its AI Assistants
Training an AI assistant requires thousands of example conversations, but real user data is hard to collect and raises privacy concerns. Adobe's…
Get the weekly digest
One Big Tech patent every Sunday. Free.