Does a machine have rights? Can you own property in a virtual world? If an AI machine paints a picture, should it get royalties?
These are just a few of the burning questions that keep AUT alumnus Sam Ennor awake at night. The selfconfessed nerd is deeply fascinated by the legal ramifications of the rapidly approaching AI revolution.
And now, armed with an AUT law degree and a decade of IT industry experience, he’s turning his soft spot for sci-fi into an extraordinary career.
A solicitor at Hudson Gavin Martin – a firm with a lion’s share of New Zealand’s tech-focused clients – Sam now gets to spend his days, as well as his nights, perched at the intersection of law and technology. It’s basically his dream job.
“While I was at AUT we were studying a case involving quite specific technology and the judge had no idea of what he was ruling on. I thought to myself, ‘hang on, that could be a thing’, and decided there was probably a niche for a lawyer with an IT background,” says Sam, who achieved the second best overall academic record in his year and made it onto the AUT Dean’s List two years running.
After being selected for the Hudson Gavin Martin summer programme in 2018, and later being offered a permanent role there, his future was assured.
“At Hudson Gavin Martin I’ve found not just one department or a couple of partners that are interested in the same stuff as me, but an entire firm that’s dedicated to it. We do so much work here that I’m interested in, and even with the more general law work there’s usually something thought-provoking and tech-related underpinning it. Plus, whenever someone gets something AI related, they loop me in because they know it’s the area I’m particularly passionate about.”
But when Sam talks about AI, he’s not just talking Facebook algorithms or smart phone voice recognition. He’s talking futuristic, space-age type stuff.
“We are in the middle of the AI revolution right now. We have deep learning AI which starts with the construction of an artificial neural network modelled on a human brain, and then uses massive amounts of data to train itself. We have creative machines which can write genuinely moving poetry without human input, and machines that can make their own decisions.
“Traditional tech has relied on people to do the work to bring on the advances, but there’s a good chance that when AI achieves a sufficient level of sophistication it will start modifying itself - that’s when we could potentially lose control. The issue with AI is that when it really takes off it will be much faster, and make much more of a difference, than we expect.”
Already there are numerous examples of people adopting the technology too hastily.
Recent research brought to light the fact that facial recognition technology was very good at differentiating people with light skin tones but struggled to differentiate people with darker skin tones, because the AIs had been fed a great deal more data on Caucasian faces.
“Obviously if this technology is used by law enforcement agencies it could lead to all sorts of issues, but luckily somebody discovered the problem and now they’re conducting in-depth research to try and address it.”
The issue of AI-caused harm is another mind bender, and one that Sam has devoted a great deal of time to. “My Bachelor of Laws (Honours) dissertation examined liability for harm caused by deep learning AI. If a machine makes a decision that leads to loss or damage, should it be liable? Usually there’s a clear chain to trace back to a human – but when you get too far into AI you end up with a grey area, because essentially you have a manmade thing making its own decisions.
“My proposed solution was to create a legislative framework providing collective insurance along the lines of ACC – I called it AICC! I thought those owning, operating or benefitting from AI machines could pay some sort of levy, which would unlock the benefits but also recover some of the liability of unpredictable behaviour.”
He admits though, it’s probably a long way off. “The issue is that technology is moving so fast that if we were to draft something new, by the time it moves through the system, the landscape will have already changed.”
But Sam – “an eternal optimist” – remains confident that one day he’ll get his day in court. “It’s my dream that by the time an AI in New Zealand causes enough damage for it to go to court, I’ll be on one side of the argument. My dissertation tackled the theoretical part of it but I’d like to see what actually happens when the rubber hits the road.”
This story was originally published in Insight, the magazine for AUT alumni and friends. Read the most recent issue of Insight for more stories of groundbreaking research and great AUT graduates who are making a difference around the world.