Technology Topic
AI and Machine Learning Patent Classification
The core AI subclass is G06N: computing arrangements based on specific computational models. Machine learning as such sits in its main group G06N 20/00, neural networks in G06N 3/00.
Computer vision got its own subclass G06V in 2022, split out of the former G06K. Applied AI additionally lands in the domain of its application, from medical diagnostics to autonomous driving.
The classification codes that matter
COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
Computational models: machine learning (20/00), neural networks (3/00), probabilistic and symbolic AI.
2018 → 2023: 37,098 → 139,015 filings (+275 %)
IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
Image and video recognition: the computer-vision subclass, established in 2022.
2018 → 2023: 18,050 → 101,211 filings (+461 %)
Who files here: top applicants
- 1 SAMSUNG ELECTRONICS COMPANY (KR) 66,669
- 2 IBM (INTERNATIONAL BUSINESS MACHINES CORPORATION) (US) 62,313
- 3 HUAWEI TECHNOLOGIES COMPANY (CN) 46,584
- 4 MICROSOFT TECHNOLOGY LICENSING (US) 41,918
- 5 GOOGLE (US) 32,969
IPC filings 2013–2023, aggregated across the codes above. Source: EPO PATSTAT.
Not sure which code fits?
Describe your technology in plain language and get matching IPC and CPC codes: