The government’s consultation on copyright and generative AI closed last week. The submissions are not yet public, but I am pleased to post my submission, which focused on an exception for text and data mining, the inclusion of copyrighted works in large language models, and the copyright implications of outputs from generative AI systems. My submission noted that the consultation raises several questions related to generative AI and copyright. I focused on three:
(1) Should Canada proceed with a text and data mining exception as recommended in the 2019 Copyright Act review?
(2) Should Canada introduce legislative reforms to address the use of copyright works in large language models (LLMs) that are central to the development of generative AI technologies?
(3) Should Canada introduce legislative reforms to address copyright-related questions arising from the outputs of generative AI systems?
My submission argues as follows:
1. Introducing a text and data mining exception into Canadian copyright law is long overdue and should proceed as a copyright reform priority. Similar provisions are widely used in other jurisdictions. Proceeding with the exception would ensure that Canada implements a copyright framework for AI that encourages innovation and investment while also providing appropriate protections for creators.
2. It is premature to introduce legislative reforms on the use of copyright works within LLMs. While there are both technical and copyright related issues related to LLMs, the copyright issue is currently before courts around the world in multiple cases that raise questions related to inclusion of copyrighted works within LLMs, whether such use constitutes infringement, and the potential application of limitations and exceptions. Given that these issues should become clearer as those cases progress, the government should maintain a watching brief to determine how the cases before the courts develop, whether market-based licensing alternatives emerge, and how the technology adjusts to reflect copyright-related concerns.
3. It is similarly premature to introduce legislative reforms to address the outputs of generative AI systems. While many have expressed concerns about the occasional similarities between generative AI outputs and copyrighted works that may have been included within LLMs, a deeper dive into the technology suggests that infringement is very rare. The courts will again be called upon to examine these claims and the government should await those outcomes before proceeding with any potential legislative reforms.
I also noted that the next statutorily mandated Copyright Act review is due. Before proceeding with any reforms, it would be useful to conduct an assessment of the implementation of the recommendations from the last review conducted by the Standing Committee on Industry, Science and Technology and scope out a future review to update on outstanding issues and address emerging ones such as generative AI.
The idea of using current texts to train a LLM makes a lot of sense. Languages continually evolve, with specific words changing the common meaning over time and between cultures.
For instance, one word (in a couple of forms) that I can think of has at least three very different meanings. In one case it refers to a bundle of wood bundled together as fuel, and in another use it is an offensive term for a gay man (both meanings from the online Oxford dictionary). In addition, in the UK I have heard it used to refer to a cigarette. The latter is a more recent meaning, so training on works that copyright has expired on is more likely to train for the former meaning.
Overall it would be good to see, rather than a completely “made in Canada” set of regulations, a basic set of regulations that the industrialized countries agree to so that the regulations are similar across the countries where these industries are expected to operate. It provides for a more consistent experience for the customers of the companies, allows the companies to use fewer models, and for the government it allows them to back up the regulations that they do put in place rather than them appearing to be more “stream of consciousness”.
Note that the above does not mean simply adopting a regulation regime that is identical to that of another country (sort of in the way that the US has been attempting to push its own regime on other countries), rather it allows for some minor local changes to address local needs (but not necessarily desires).
Things got out of order in my editing of the comment. In the second para when I am referring to “the latter” it is the meaning which refers to a gay man.
Being unemployed is no fun. Google pays $167 in Hr USD. To help keep the money flowing while you look for a new job, hn governments offer unemployment protection to unemployed workers.
Go here…….. https://GooglePayMoney76.blogspot.com
I will be returning to this blog often; I have bookmarked it.
Pingback: Law and Media Round Up – 29 January 2024 – Inforrm's Blog
Your analysis on the impact of generative AI on Canadian copyright law is quite insightful, Michael. The idea of introducing a text and data mining exception is particularly intriguing. How do you envision this exception striking a balance between the rights of copyright holders and the imperative to foster innovation in AI? Could you point to any models or examples from other jurisdictions that Canada might look to for guidance?
Additionally, I’m concerned about the security implications associated with the advancement of artificial intelligence. With the rapid pace of AI development, how do you think we should address the potential risks this technology poses to privacy, security, and ethical standards? Are there any safeguards or regulatory frameworks you believe should be implemented to mitigate these risks while still encouraging innovation in the field?
References:
https://www.forbes.com/sites/waynerash/2024/02/07/generative-ai-exposes-users-to-new-security-risks/?sh=590ef3ef2dfe
https://www.spiceworks.com/tech/artificial-intelligence/guest-article/tackling-security-risk-of-ai-llms/
https://socanadiancasino.com/casino-bonuses/free-spins-bonus/
https://www.theinformation.com/articles/why-were-worried-about-the-wrong-ai-security-risk-gemini-is-coming-again
Hack] Unique A.I. App Makes Us $635/Day
Well. That wasn’t unexpected Furnace Contractor Service
FlipBooks are a great addition
to any passive income strategy. Because once you create a FlipBook, market it, share it & Earn it, it can technically sell itself.
Learn More https://www.youtube.com/watch?v=A9ai7ey85dE
16764
fine/
The administration should wait for the courts to review these claims once more before moving forward with any possible legislative revisions.