Artificial Intelligence Flashcards
AI system according to OECD recommendations (2019)
a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments. AI systems are designed to operate with varying levels of autonomy
AI
powerful algorithms (super software) acquiring fundamental human-like capabilities, such as vision, speech, navigation, etc.
Machine Learning
▪ a technique supporting AI
▪ consists of methods that allow computers to draw conclusions from data and provide them to AI applications
Deep Learning
▪ a technique used for implementing machine learning
▪ teaches computers to learn by example and using examples for self-thinking
No uniform, binding legal definition of AI under IP/patent law and no specific AI-focused patent regulations
2017 had UE Resolution on Civil Law Rules on Robotics
(2015/2103(INL)
EC Communications: Artificial Intelligence for
Europe and Building Trust in Human Centric
Artificial Intelligence with ethics guidelines for
trustworthy AI (COM(2019)168)
2020 had WIPO issue revised guidelines
Issues surrounding AI as a subject of protection
AI as a new category of an invention?
‚Legalization’ of patentability of computer programs?
Sensitive AI- inventions as a subject of a new exclusion
from patentability?
Patentability of AI under Article 52 EPC
exclusion from patentability of „schemes, rules and methods for performing mental acts, playing games
or doing business, and programs for computers” (but only to the extent to which a European patent application or European patent relates to such subject-matter or activities as such)
Patentability of AI in the US
Exclusion from patentability of patent claims that are directed to abstract ideas (e.g. a mathematical
algorithm) - exclusion of patent claims directed to subject matter that could be performed through an “ordinary mental process”, “in the human mind” with the limited exception for claims that specifically provide ways to achieve technological improvements over the tasks previously performed by people
AI core invention
inventions characterized by mathematical or statistical information processing technology that forms the basis of AI, such as various machine learning methods including neural network, deep learning etc.
AI applied invention
inventions characterized by applying AI core invention to various technical fields such as image processing, speech processing, natural language processing, device control/robotics, various diagnosis / detection / prediction / optimization system, etc
AI assisted invention
when human inventor uses AI as a tool to invent
AI implemented invention
where invention is developed by human but a claimed technical solution (application in technical field)
is achieved through AI component of a claimed
subject matter
AI generated invention
where AI generate an invention autonomously with little or without human control
EPO approach on AI inventorship
➢ an inventor designated in the application must be a
human being, not a machine
➢ a concept of an ‚electronic personality’ of the inventor should be refused
➢ the name of a machine does not meet the requirements of Rule 19(1) EPC (the requirement of a designation of the inventor not met)
Opportunities for patent offices with AI
▪ integration AI and machine learning into examination systems to help streamline the patent application
process - application of AI to auto-classify patents and connect them to the correct examiners more rapidly
▪ automated help for the research performed by staff examiners and supervisors
- „virtual assistants” extracting information from prior art data bases
- artificial neural network technology to translate patent dossiers (e.g. Machine Translation through
Google’s Neural Network)
- automatic annotation of patent literature: indexing numerical values, chemical compounds
