Share with your friends

Building trust in autonomous vehicles: The AI and ethics challenge

Autonomous vehicles: The AI and ethics challenge

With the challenges of increased population, urban density and changing consumer expectations, better use of data and the power of artificial intelligence has the potential to transform our infrastructure and how we meet the needs of our communities. However, it is critical that we identify and address a range of issues associated with privacy, ethics and automated decision-making.


Also on

With the release of KPMG’s Autonomous Vehicle Readiness Index 2019, Adrian Box, KPMG’s Chief Data Officer for Global Infrastructure, caught up with Kate Marshall, a Partner in KPMG Law focusing on privacy, technology and data and David Evans Head of Data Science for KPMG, to discuss some of the opportunities and challenges.

Kate Marshall on ethics and privacy

AB: Data and AI are creating a lot of conversation, both positive and negative. This is particularly the case in relation to autonomous vehicles. What do you think needs to change for the Australian community to feel more confident about autonomous vehicles?

KM: Well I think it is interesting, we want to create an environment where there is trust. At the moment the debate is focused on whether the vehicles themselves are going to be safe. Assuming we can solve that, I think the more interesting conversation we should be having is around the decisions that are in built into the algorithms that they embed in those vehicles.

Who’s going to make that decision around whether the vehicle swerves to hit your cat, or potentially the vehicle next door, or the pedestrian? Who is going to make those ethical decisions and how can we be transparent around what those decisions are, so that we can build understanding, transparency and trust within the community around how these vehicles are actually going to operate.

AB: Do you think we have the right legal structures in place around the use of AI?

KM: I think there is a couple of layers to that.

There is the simplistic issue that we need to change the legislative environment for autonomous vehicles when you no longer have a driver. That’s being addressed. All of the transport ministers are looking at that together with the National Transport Commission.

The next consideration needs to be around the use of the AI that will come into those vehicles and how to create the regulation or at least the transparency around how those algorithms impact the vehicles. What are ethics behind them, what’s the bias within them, do we understand how they are going to work, who has embedded those decisions and are they in line with community expectations?

This is where KPMG is getting involved. We are hosting an AI and Ethics Forum, bringing together industry leaders, academics and regulators to talk about these questions. I think it is a really important conversation.

AB: Are there unique privacy issues associated with autonomous vehicles?

KM: Not particularly. In a lot of ways it is very similar to ride sharing services that we have all become comfortable using. Yes, there is going to be huge amount of data created by autonomous vehicles. There is also going to be a need and a desire to share that data and that’s where there is going to be some decisions that need to be made. Some of that data needs to go back to regulators or enforcement authorities to cope with the parking fines and when there has been an accident. Clearly that data needs to be shared. Otherwise, I think customers or individual passengers want to have some say in what information about themselves will be shared and I think our privacy regime allows for those opt-in / opt-out mechanisms to provide those controls.

David Evans on AI and the adoption of new technology

AB: Where do you see AI adding the most value in the infrastructure space?

DE: Infrastructure is a very broad area covering everything from hospitals to schools, utilities and transport but where I see the most value is in transport. In particular, safety and autonomous vehicles come to mind. It is quite topical, but where I see great value is things like tollways and freeways. Where you have got traffic congestion, you have got algorithms looking at the flow of traffic and making adjustments to things like speed and things like that, based on patterns of traffic flow that may be emerging.

AB: Where do we draw the line on giving algorithms/bots the freedom to learn autonomously?

DE: I don’t think there is a hard and fast line, but more to the point we need to have controls and guardrails in place for situations where algorithms are incrementally learning. Traditionally a lot of algorithms actually have what’s known as batch learnings, so they are pre-configured with patterns they have learnt. Whereas things like autonomous vehicles and transport you made need to actually take information and learning new patterns. Machines and algorithms are very good at doing a very specific task, whereas humans can learn new things and not forget old things. There is still a lot of research and work to be done with getting machines and algorithms to learn on the fly. I don’t think there is a hard and fast line it’s just a matter of getting those controls and guardrails in place.

AB: What do you see as the biggest challenges of AI in terms of autonomous vehicles?

DE: In terms of the challenges for autonomous vehicles, it’s perception and uptake by people. As this technology becomes more advanced, people will become more comfortable and it depends very much on the context. For example you have got cars, but you have also got aircraft with different consequences of things going wrong. It is having that level of comfort in the adoption of new technology.

AB: Where else could AI support infrastructure?

DE: Another great example is utilities, such as gas or water or electricity. You may have tens of thousands of power poles and other substation-type infrastructure, where at the moment you have got people going around and manually inspecting the assets. There is now people working on drone technology where the drone flies over the assets and takes photos of different components. They can recognise which ones require upgrades or maintenance and reports that back to a central location. You can imagine the millions of dollars in savings that are available.


As you can see, data, AI, autonomous vehicles and other technologies provide us with fabulous opportunities to create and support thriving communities. However, they also create significant new challenges. A detailed understanding, active debate and real community engagement, will be critical to us successfully leveraging these technologies into our daily lives. Without this, trust will not be achieved, or will be broken, which will impact on the speed and effectiveness of incorporation.

©2021 KPMG, an Australian partnership and a member firm of the KPMG global organisation of independent member firms affiliated with KPMG International Limited, a private English company limited by guarantee. All rights reserved. The KPMG name and logo are trademarks used under license by the independent member firms of the KPMG global organisation.

Liability limited by a scheme approved under Professional Standards Legislation.

For more detail about the structure of the KPMG global organisation please visit

Connect with us


Want to do business with KPMG?


loading image Request for proposal

Save, Curate and Share

Save what resonates, curate a library of information, and share content with your network of contacts.

Sign up today