Legal Affairs for Big Data and Artificial Intelligence (AI)
Ever since information about AI and the law and AI and intellectual property rights was first disseminated,
STORIA Law Office has been receiving inquiries from numerous business enterprises involved with AI matters
about legal consultations.
For example, these legal consultations have included such issues as whether a trained model using raw image
data on the Web can be generated without the consent [of those whose images are used]; whether it is allowed
to generate a trained model based on a dataset created in violation of terms of a license; what the best way
is to develop business while still legally protecting the generated trained model; and who is liable, if by
chance, there is an accident involving generated AI.
Although AI has tremendous potential, discussions about how to approach such legal aspects as AI’s
generation, protection, and use are still in the early stages.
STORIA Law Office provides legal consultations about the following 4 AI-related field:
1. Generation of AI
2. Protection of AI
3. Use of AI
4. Legal liability arising from the use of AI
Big Data: Lawful collection and use of Big Data
Although some Big Data contains certain data that belongs to no one such as sensor data, other Big Data
contains personal information.
When using such Big Data containing personal information in business, it is necessary to obtain consent to such
use or employ various devices, such as processing measures to anonymize the data.
Big Data: Negotiation for Contract Execution and Drafting of Contracts
With respect to Big Data, since there are no uniform rules concerning who owns the right to use the data, there
is also no definitive, conventional wisdom about what issues should be addressed in a contract or what
contractual terms should be included. On the other hand, there are such guidelines concerning the right to use
Big Data as the “Contract Guidelines on Data Transactions” (October 2015, Ministry of Economy, Trade and
Industry) and the “Contract Guidelines on Data Utilization Rights ver. 1.0” (May 2017, IoT Acceleration
Consortium, Ministry of Economy Trade and Industry).
At STORIA, we provide support in drafting of contracts and negotiating for contract execution involving Big
Data based on a clear understanding of these various guidelines.
AI: Legal Issues related to Generating AI
In the generation phase of AI (trained model), various actions are taken, such as collecting, creating a
database, duplicating, adapting, machine learning and deep learning, which involve raw data, databases, training
datasets and trained models.
STORIA has provided numerous legal consultations concerning in what situations these actions can be lawfully
taken and, on the other hand, in what situations these actions would be unlawful.
AI: Legal Issues related to Protecting AI
AI (a trained model) itself generated by an appropriate method based on a large voluminous dataset processed
with an effective processing method has extremely significant intrinsic value.
In addition, these trained models can generate even further learning-enhanced second and third generation
[trained models] and can be transferred to other business areas.
On the other hand, if these trained models are not completely legally protected, the incentive to invest in AI
will decrease.
Technology, contracts and laws (the Patent Act, the Copyright Act, and the Unfair Competition Prevention Act
(trade secrets)) can be used as methods for protecting AI trained models.
At STORIA, we use a suitable combination of these methods to provide detailed advice and a practical solution
to protect your trained model, one of your valuable corporate assets.
AI: Legal Issues related to the Use of AI
When various content is generated by the use of generated AI, who owns what kind of rights to the contents?
Actually, this has been a problem for a long time.
In a situation where content that exceeds human creation capability is increasing exponentially, STORIA can
advise you about how to protect content generated using AI.
AI: Legal Liability Arising from the Use of AI
The inside of AI is a black box, and humans cannot understand what kinds of processes are performed inside.
However, since AI is obviously not versatile (at least at this point), it is a matter of course that some
accident through the use of AI may occur.
If such an accident occurs, who has, or does not have, liability? There is no settled opinion about this issue.
Therefore, although there are many difficult situations in our legal consultations at this point where we
cannot give a definite answer indicating that “If you do this, then you will have no problems”, we can advise
you about a strategy for maximum risk reduction.
