I’ve recently been involved in several presentations and expert panels on the benefits, challenges and social consequences arising from a future alongside robots, artificial intelligence (AI) and associated technologies.
My involvement recently at the International Joint Conference on Artificial Intelligence (IJCAI) reinforced the reality of a new paradigm that could shake the very foundations of our concept of intellectual property (IP) ownership.
We have longstanding laws on how, when and where we use machines such as vehicles and airplanes. But what about machines that mimic the brain or have some form of intelligence? This is a new proposition.
As businesses increasingly invest in new AI and robotics technologies, engaging in research and applying them in innovative ways to enhance competitiveness, we are seeing more examples of original works created not by humans, but by autonomous AI.
However, investors should tread with caution while questions remain as to the ownership of such works.
Can a robot ‘own’ work?
Who owns intangible outputs which could be perceived as IP when they are generated by a robot or artificial intelligence (AI)? Who owns the IP – the manufacturer, developer or programmer? Could ownership fall to the user providing the data for the robot to create the output? Or alternatively, could the robot own its creations?
Our regulatory framework generally assumes that ownership of IP works pertains to natural or legal persons. However, what happens when the source code, objects or other assets are created autonomously and are directed by non-human entities, as will increasingly be the case in the future?
Creativity may no longer be the exclusive province of humans.
The Japanese business of McCann advertising has a Robot-AI creative director working on commercial campaigns while AIs have also been responsible for generating original and commercially successful music, paintings, writing and even sculpture.
The 2010 Emily Howell CD “From Darkness, Light” represented a milestone in AI music composition. Emily Howell is an AI built by David Cope which uses a rules-based music writing program along with an algorithm to insert some randomness into the musical pieces. While the AI draws on a huge library of musical influences, as well as mentoring from David himself, he claims the software is its own “musical master”.
In the visual arts world, a robot artist called Aaron has been developed by Harold Cohen, who worked for some time as a guest scholar at the Stanford University AI Lab. Aaron is completely autonomous and even washes his own brushes, like any human artist. Aaron’s works have reportedly hung in major galleries around the world.
Similarly, AI systems are being used by major publishing companies to write news reports on sports matches as well as to generate financial news, corporate earnings previews and other data-intensive articles.
Given that technological research and progress are often driven by the promise of financial rewards, this uncertainty around ownership could be a disincentive for commercial entities to invest in AI development.
Are robots ‘electronic persons’?
In February this year, the EU Parliament, with an unprecedented show of support (396 for; 123 against; 85 abstentions), took an initial step towards enacting the world’s first Robot laws.
While the Parliament’s motion is yet to be ratified by the EU Commission, the decision is certainly generating considerable discussion as it calls for sophisticated autonomous robots to be given specific legal status as electronic persons. This could potentially allow them to own their creations, as well as being liable for any damage they cause.
The UK had previously implemented specific provisions to protect literary, dramatic, musical or artistic work which is computer-generated. The UK Copyright Designs and Patents Act 1988 defines ‘computer-generated work’ to mean work ‘generated by computer in circumstances such that there is no human author of the work’. The author is the person who undertook the arrangements necessary for the creation of the work.
It’s unlikely the UK Government was thinking ahead to AI and robots when it passed this law, but since these technologies are advanced manifestations of computing, the provisions are at least reflecting and in alignment with the digital age.
Despite recommendations to adopt the UK approach by independent Copyright Law Review Committees over a period of 10 years, the Australian Government is yet to act. This has resulted in a situation where the wider use of technologies like AI will hinder the securing of copyright protection in computer-generated or AI-generated works.
I was disappointed to see that the neither the Australian Productivity Commission Review of intellectual property arrangements, nor the Federal Government response, made any mention of AI or any autonomous technologies.
Are we keeping up with the realities of our times and the consequences flowing on from rapidly developing technologies like AI?
Under existing Australian law, any works created by autonomous AI and robots will suffer serious hurdles in securing copyright protection and might not have sufficient ‘authorial’ contribution for copyright to subsist.
The pace of technology evolution will only increase and it’s challenging for our regulatory system to keep up. And yet, we must if we are to maintain an environment that encourages innovation, sound investment and growth.