Tasmanian Parliament AI Data Centre Inquiry Hearing – 4 November 2026
CHAIR – Thank you both for appearing at today’s hearing. If you could please state your name and the capacity in which you are appearing before the committee.
Mr COLEMAN – Thank you. Luke Coleman, I’m the CEO of the Australian Telecommunications Alliance.
Ms ROBERTSON – I’m Victoria Robertson. I’m the Head of Government Relations and Policy at the ATA.
CHAIR – Thank you, and can I confirm that you’ve received and read the guide sent to you by the committee secretary?
Witnesses – Yes, we have. Thank you.
CHAIR – This hearing is covered by parliamentary privilege, allowing individuals to speak with freedom, without fear of being sued or questioned in any court or place out of parliament. This protection is not accorded to you if statements that may be defamatory are repeated or referred to by you outside the parliamentary proceedings. The hearing is public. The public and media may be present. Should you wish aspects of your evidence to be heard in private, you must make this request to the committee at this time.
I’ll just quickly do introductions. I’m Brian Mitchell, the Chair.
Ms BADGER – I’m Tabatha Badger, the Deputy Chair, member for Lyons.
Mr SHELTON – Mark Shelton, member for Lyons.
Ms OGILVIE – Madeleine Ogilvie. Hi, how are you?
Ms BURNET – Helen Burnet, member for Clark
Prof RAZAY – George Razay, member for Bass
CHAIR – And online we have?
Ms DOW – Anita Dow, member for Braddon.
CHAIR – Thank you, members. Mr Coleman, Ms Robertson, would you like to make a quick statement?
Mr COLEMAN – Yes, that would be great, thank you. Thank you very much, Chair, and members of the committee for the opportunity to appear. The ATA is the peak body for the Australian telecommunications industry. We represent network operators, technology vendors, service providers, digital infrastructure operators and ICT service providers – so, the full gamut of the telecommunications sector in Australia.
Our submission has one very simple proposal: that Tasmania should seek to capture the broader economic and infrastructure benefits of AI investment and not simply provide a location for data centres. Now, that means attracting responsible investment while ensuring that projects are properly planned, meet the reasonable costs that they create and deliver tangible benefits for Tasmanians.
Our particular focus is on one of the critical enablers of those benefits, and that is the telecommunications infrastructure that is needed to connect data centre facilities.
So, we want to emphasise three points:
First, data centres provide computing capability, but it is telecommunications networks that make that capability accessible and useful. Networks connect AI capability to businesses, governments, researchers and, of course, communities, and without sufficient high-capacity connectivity, the benefits of data centres really can’t extend beyond the data centres themselves.
Second, telecommunications infrastructure must be planned alongside data centres. Land, power and water are not enough to make data centres ready to operate. Before major facilities are approved, there needs to be a clear understanding of the telecom’s capacity that it will need, the fibre routes going into the site, the entries that are required, and the necessary approvals and land access required and how that gets secured. So, data centre approval alone does not itself approve the fibre routes or the civil works that are required to get that site connected to telecoms infrastructure. If those requirements are only considered later, the result can be delays, additional costs, and a less efficient or less resilient design.
That’s why the ATA, on behalf of our telco members, is recommending an integrated infrastructure and connectivity plan that would come as part of the initial development application and not as a condition dealt with after the approval. So, the facility and it’s enabling infrastructure should be assessed and sequenced together.
The third point we make is that resilient connectivity requires physical route diversity, not simply multiple service contracts. Two services are not necessarily two resilient connections if they rely on the same conduit or the same crossing or the same exchange, or any other common point of failure. In some cases, genuine resilience will require new and physically separated fibre routes. Tasmania’s planning framework should recognise that need and provide a proportionate pathway for strategically important underground telecoms infrastructure while retaining targeted environmental, cultural heritage and land-use safeguards.
This is the practical deployment challenge. A fibre route only works when every part of the route can be delivered.
I’ll conclude on this – that a delay affecting one short segment, crossing or land-access approval can hold up the entire route. So, removing one bottleneck will have limited effect if the next approval constraint remains unresolved. The central recommendation, then – I’ll finish my remarks on this – is that we need to plan, assess and approve the data centre and the telecommunications infrastructure as one integrated system. That means engaging early with telecommunications providers, coordinating approvals processes for telecoms infrastructure, and protecting the corridors and route options needed for future capacity as well as resilience. That approach would help Tasmania identify any constraints early, allocate project-specific costs appropriately, avoid unnecessary duplication, and ensure that data centre investments deliver reliable capability and lasting benefits for the Tasmanian community. Thank you, Chair, and over to the committee for any questions.
CHAIR – Thank you, I’ll kick off. I was going to ask you what your three most critical issues are, but I think you’ve covered that off in your intro. Thank you for raising the issue of telecommunication infrastructure, because it hasn’t been largely touched on. We’ve had a lot of talk about power and water; you’re one of the few submitters to talk about the need for the superhighways of connectivity that are required. The question I’m going ask you is: who pays? Under your scenario for infrastructure planning, pre-approvals and precinct planning, who pays?
Mr COLEMAN – The telecommunications sector has a real appetite to be investing in this infrastructure. We are not here looking for handouts; we are not here looking for government subsidies. Telcos want to be building this infrastructure to data centres. All we are asking for is reforms to planning rules to make that as streamlined as possible.
We see with data centres around Australia – data centre operators aren’t just looking for geographic diversity; they often talk about ‘quadversity’. They want at least four geographically separate entry points into data centres; it can even be higher than that. The reason behind that is there is zero appetite for downtime at all; they want 100 per cent uptime in these data centres. If you have one fibre cut, two fibre cuts, even three fibre cuts; if you’ve got four separate fibre routes going in, you can still keep that connectivity enabled. But to enable that, to enable at least four separate routes, we need to make it a lot easier to deploy that infrastructure.
Sorry to deviate a bit from your original question, but the answer is: it is the telco sector itself which has this incentive to invest. There are well-established commercial relationships with both the data centre operators themselves, as well as the tenants of those data centres, which are major over-the-top technology platforms. We really just want to be working with state and Commonwealth governments to ensure that planning approvals aren’t the bottleneck in getting that fibre delivered.
CHAIR – Sorry, I don’t want to dominate, I just want to be very clear that – so what you’re saying is you see state and federal, and some degree local, governments enabling and facilitating through planning. But you see the commercial, the bills being paid, by the operators and the tenants as a commercial arrangement between them and your members. That’s essentially the guts of it?
Mr COLEMAN – That’s right. This is a purely market-led play. There is a very strong investment appetite out there. The constraint that we see in the industry is not an investment constraint, it is a planning constraint, and that’s for getting in those major fibre-optic cables. It’s not only fibre-optic cables, it’s also mobile infrastructure. Remember that for many Tasmanians that are going to be using AI services, they’re accessing it on the go, and so having more mobile tower infrastructure built in the state is also a real outcome we should be aiming for here; so that consumers can access the services that are hosted in those data centres. But primarily what we’re talking about when it comes to planning reform is making it as easy as possible to get that fibre in the ground and into those data centres.
CHAIR – Thank you.
Ms BADGER – You’ve touched a little bit onto where my question is going to go in terms of how Tasmanians might use this. It’s kind of in two aspects. I guess, is there any jurisdiction, particularly in Australia, that is looking well in terms of the long-term planning of getting this infrastructure underway already – notwithstanding, of course, the AI infrastructure boom, if you like, has taken everyone off guard to a degree at scale? So, if there’s anywhere that we could sort of look to that is doing it well? And then what restraints you see, perhaps, in addition to just planning in Tasmania, that we need to catch up on, on this front?
Mr COLEMAN – Sure. I can certainly go to the second part about constraints in Tasmania. We do have a few examples. I might speak to your first question just generally at first, and then I’ll turn to my colleague, Victoria, if she wants to add any specific examples that come to mind.
The states that are doing this best are the ones that are seeking to harmonise across various government departments’ planning rules. And when I say government departments, we’re talking about state, but particularly local. We’ve got more than 550 councils around Australia. Each of them have overlapping, duplicative or unique planning approval rules that make it very challenging for an infrastructure investor to know if they’re going to get approved, how long it’s going to get approved. And so the states that are doing the best job here are the ones that are harmonising council approval processes. That really cuts down on the approval time to get fibre in the ground, or to get mobile towers built.
And so, we’re seeing across Australia there is a real variance between states. There are those that are embracing this opportunity to align planning rules across councils, and they’re the ones where we’re seeing the most traction. There are other states that aren’t really tackling that problem, and the real-world examples we’re seeing from that is operators that want to build big inter-capital fibre routes get bounced from one department to another. They have one issue that’s in the department of environment. Then, maybe that piece of fibre crosses a river, they get bounced over to the department of water or fisheries. Perhaps there’s some local vegetation which is an issue, so they get bounced back to another department to deal with that particular issue.
The more we can have government working in harmony – and this was one of the key recommendations of our submission – to have a coordinated approach. You might call that a coordinator-general or a similar role that is there to make it as simple as possible to roll those out. So, why don’t I pause there for a moment? Victoria, did you want to add anything on that front?
Ms ROBERTSON – Ms Badger, you asked about any particular states that could provide examples. In terms of the fibre rollout, it’s worth having a look at some of the new changes that have happened with the Victorian government in relation to rollout of fibre. I don’t have the names of those programs or anything in front of me at the moment, but happy to take that on notice to provide that later.
Ms BADGER – Thank you.
Mr COLEMAN – I might just add to your second question. You’re asking about specific bottlenecks in Tasmania. I’d draw your attention to the appendix in our submission where we’ve called out one, two, three, four specific changes that we would really recommend the Tasmanian government investigate.
So, in Tasmania you have the Tasmanian Planning Scheme. Within that planning scheme you’ve got the telecommunications code that deals with the rollout of telecoms infrastructure. We’ve recommended a few very minor tweaks to that telecoms code that really take a more risk-based approach to planning rules around telecoms infrastructure.
One of the big issues that we wrestle with Australia-wide is that a particular piece of telecoms infrastructure gets assessed – for the whole thing it might be an inter-capital fibre route. But there will be specific areas that need to be focused on. Maybe there’s an environmental impact area – that you might have an endangered species, just as an example. Can we make that –
CHAIR – Mr Coleman? I’m just going to, sorry, it’s the Chair. I’m just, would you mind just trying to bring your answers down a bit. We’re running out of time, otherwise.
Mr COLEMAN – That’s fine. I’ll be a bit more succinct. Apologies. That we want to just take a risk-based approach; look at the areas where there is real sensitivity that we need to address, and focus the attention of the approvals process on that, rather than the entire project. But I won’t speak to it any further. You can have a look at the appendix in our submission. Thank you.
CHAIR – Thank you, Ms Ogilvie.
Ms OGILVIE – Yeah, thank you very much. I’m interested in the telco side of things as well. Thank you so much for joining us. For over a decade, I’ve made it a bit of a mission, personal mission, to try and get us a new subsea cable. It’s been very difficult. We now have an opportunity, where the market is prepared to move on that. I’m just wondering whether you could explain to us, in whatever terms you choose, the importance of resilience and also the latency issue with the new subsea cable, and the potential economic uplift that you see happening when new high-speed cables go in, particularly as an island state. We’re very interested in this.
Mr COLEMAN – Yeah, I think that’s a fantastic outcome for Tasmania if we get another submarine cable coming in. Resilience, as you mentioned, a great point that. You might have four fibres going into the data centre but if, ultimately, your single point of failure is one submarine cable going across to the state, well, that is a huge risk and so having another cable in place makes a big difference, as well as making a big difference economically. That becomes both a resilience bottleneck, but also an economic bottleneck if you’ve only got a monopoly provider there.
Really, while we’re talking about submarine cables for Tasmania, this is just another intercity fibre route, and this is where the investment is coming in telecommunications. It is between major data centre hubs, which are all in capital cities, so we’re seeing huge investment going to Sydney, Canberra, Melbourne, and if we want to attract data centres into Tasmania, having that intercity connectivity coming through to Tasmania as well, in your case, you’re unique that it is a submarine cable, not just another intercity fibre route but it is absolutely fundamental infrastructure to enable data centres.
Ms BURNET – Thanks for your detailed submission. In 6.2 you talk about standardised and meaningful resource disclosure. Can you just talk about why it’s so important to have good public reporting of the material resource consumption?
Mr COLEMAN – Sure, I might call on Victoria for that one.
Ms ROBERTSON – Sure. Sorry, I’m just going back to the submission as well just to make sure that I’m answering it consistently. Look, so basically, I think what we’ve said in the submission is that having standardised information means that each project, each investment can be assessed on the same merits. It’s the same criteria; everybody’s being assessed the same. It’s almost like a level playing field. So, basically, it will – you’ll be able to, I guess, assess the actual impact of a project as well, if you’re measuring every project in the same way. So, that’s essentially why we’re talking about that.
Ms BURNET – And you see that as beneficial to not only proponents and government, but also for the public, knowing what’s happening?
Ms ROBERTSON – Yes, so the public discourse around data centres varies from one spectrum to the other and, at this point, getting appropriate information about projects out there with communities can actually be of a benefit to getting that project rolling out in a timely and efficient way. So, there are some parts of information that you probably don’t want to disclose to public and we have mentioned those in our submission as well and they could be around technical or points of vulnerability or other sort of commercial-in-confidence types of things around networks.
Mr COLEMAN – I might just put a small bookend on that comment, which is this comes back to the issue we see with more than 500 councils around Australia, each with their own regulations around the information that needs provided. You’ll see in section 6.3 of the submission about ‘report once, use many times’. We think this is the ideal; we don’t want these infrastructure investments to get tangled up in red tape and that really is the case around Australia. The information that is needed, the more it is standardised, the quicker you can have these projects rolled out and so you’ve got people who are actually digging up the ground rather than shuffling paperwork.
Ms DOW – Thanks very much for your presentation to the committee. I was just interested if you could expand more on your notion of whole-of-corridor planning and then you go on to speak about the integrated infrastructure and connectivity plan as part of a planning approvals process. Could you elaborate on that a bit, on those two concepts, please?
Mr COLEMAN – Sure. The issue that we often face is this, particularly with long‑haul fibre routes, so inter-capital fibre: you might have a submarine cable that’s going on to the north coast of Tasmania, maybe into Burnie. It’s got to get down to Hobart and other parts of the state. You will find, in most parts of Australia, individual sections of that fibre route are subject to individual planning approvals from not only local councils but also various different state government departments from their own jurisdictional responsibility, whether it’s environment, heritage, native title, different land access approvals.
What we’re asking for here is to treat it as one piece of infrastructure that, rather than having to seek literally thousands of different approvals from different landowners, different property types, different jurisdictions and government departments, it is treated as one piece of infrastructure, like a main highway would be, or to use the energy sector example, if you’re putting in a new major transmission link for energy. Telco is always the exception here because, unlike every other utility, we are not regulated by states like energy and water. So, the states have jurisdiction over energy and water and how they’re deployed. Telco is regulated at the Commonwealth level, but we still have to wrestle with all of those state‑based planning rules, environment rules, those sorts of things. So that’s what we’re referring to when we say: treat it as one continuous piece of infrastructure to minimise that red tape.
Ms DOW – And do you have examples of other parts of the country where that mechanism is used, or is that a new concept that you’re introducing?
Mr COLEMAN – This is definitely something that we’re wrestling with around the country. I’ll give you one example, in Western Australia, one of our members, which is a major fibre backbone network provider, was building a new fibre route through the Pilbara. It was the first competitive fibre route through the Pilbara. In that project, one single 30 kilometre stretch of that project, which was thousands of kilometres long, but just within that 30 kilometre stretch they we’re dealing with at least four different government departments and at least 12 different title holders, whether they were just landowners, native title holders, tenement title holders. All of that just adds complexity and red tape – each of those different types of title holder or government department would have different timeframes involved in consultation processes. One might have a 30-day minimum period for consultation, another might have 90 days. Sometimes they can be extended. All of those differences in one 30 kilometre stretch of a project that stretched thousands of kilometres delayed that project by more than a year and a half. That adds up to real money, to real delays for the community and getting access to that infrastructure, and that is a real opportunity for planning reform.
Prof RAZAY – With the rapid expansion of AIs – more than 160 in Australia, more than 13 million people using AI – at what stage do you start to get really concerned about the capacity of telecommunication networks in coping with this? Do you have any future modelling about how far you can cope with?
Mr COLEMAN – Sure. We do see AI as a significant driver of telecoms investment right now – first of all, in fibre backhaul capacity, international capacity and interstate routes. So certainly, the investment appetite is there. So, am I concerned about constraint? The main constraint that I’m concerned about is getting it built in time because the investment appetite is there. Getting it rolled out quickly enough, when there’s so much red tape around planning rules, that’s the biggest constraint, and I would add to that mobile capacity as well. One of the major trends that we’re seeing is AI moving more capacity onto the uplink. So, historically networks have had majority of downlink. People are downloading videos, they’re downloading content. AI means people are doing more on their devices and that information needs to get uploaded to the data centre, and so there is a change in the way networks are being designed to cope with that additional uplink requirement.
Mr SHELTON – My question ‑ and you mentioned the resilience and data centres needing one, two, three, four access points to the network. We’ve seen with our energy networks gold plating and the costs that are incurred through that. So where is the reasonable balance? How does ‑ yes, what is needed?
Mr COLEMAN – I’ll have a first crack at this. I’ll let Victoria jump in if she wants to add anything, but major difference here with energy networks. Energy networks, transmission networks in particular ‑ typically a monopoly provider. So, they’re very heavily regulated from a price perspective. Telecoms networks are competitive. You have got a number of operators that are queuing up to deliver this infrastructure to win those commercial contracts. So, gold plating is a problem that is solved by the market dynamic that you’ve got a number of operators that all want to provide that infrastructure. They all want to secure that contract with the data centre operator or the tenant in that data centre, so you don’t need the same approach to telecoms pricing as you do with energy, which is usually a state‑based monopoly provider because the market solves that problem. Vic, did you want to add anything to that thought?
Ms ROBERTSON – Just adding onto the market sort of point there as well, that when you’re talking about the gold plating of an energy network, obviously that comes at a cost to the consumer. And from what we’re seeing from the telcos and from your ‑ I guess, everybody’s mobile prices – is that it’s one of the utilities that everybody uses every day, one of the essential services people use every day where the price has actually decreased over time for the services that you get. So, the market does really balance out that need for gold plating or, you know, providing the essential service that’s actually needed.
CHAIR – I just want to come back to Ms Dow’s question about the one‑stop shop, for want of a better word, and you’ve talked about wanting to reduce red tape. I just want to come at it from the counterview and put that to you, that the benefit of all that regulation is to safeguard certain things: safeguard native title rights, safeguard biodiversity, safeguard threatened species. What mechanism do you see replacing that multiplicity of regulation that will still safeguard all those things?
Ms ROBERTSON – That’s a very good question, Chair, and what we’re sort of saying through our submission is that we’re not trying to remove those safeguards from community, environment, cultural heritage, or anything like that. We’re not asking for that. We’re asking for proportionate regulation. We’re asking for streamlined approaches to being able to roll out infrastructure so that communities actually get the benefit in a much quicker way. By providing a framework like we’ve discussed in our submission around the integrated infrastructure plan, what that means is that coordination can happen at a higher level. So, it can be at a state government level, and basically those decisions can be made in a one‑stop shop, which will speed up the ability to get the deployment rolled out and also get those benefits out to the community a lot quicker. At the same time, the state government is still safeguarding those obligations.
CHAIR – I totally get the desire for a more streamlined process. I don’t want to get into the weeds on this, but if, for example, you deal with the Office of the Coordinator‑General, who’ll be your one‑stop shop to deal with an application, the application is still bound by the requirements of the environmental protection diversity act, for example – if that’s right name for it, I can’t remember. The Coordinator‑General still has to wait upon those processes to go through before their office can deal with it. I just wonder in terms of, what are the practical, mechanical changes to processes that you see will streamline things without abandoning safeguards?
Mr COLEMAN – I think the best thing I can point you towards, again, in the appendix of our submission we go through four very specific tweaks, and I deliberately call them tweaks, because these are relatively small changes that will make a major difference. They do not remove those regulated protections, whether it’s for environment, native title, or whatever other protections are there. Rather, it seeks to focus them on areas of actual risk. And by having that done through a coordinator‑general, it will mean that you have a more streamlined approach, rather than dealing with separate agencies that have all got their own timeframes that might have conflicting rules in place. It’s something just to keep the process focused.
To quickly refer to a couple of things we’ve recommended ‑ to expand the clause in the communications code within the Tasmanian Planning Scheme that would provide exemptions for certain types of infrastructure. So, just to give you a bigger picture: this is how telco infrastructure is built all around Australia ‑ that we have in the Commonwealth law a thing called carrier powers and immunities. Most telco infrastructure is built within the bounds of those carrier powers and immunities; you know, towers can only be a certain height, facilities can only be a certain size. But those immunities can be constrained by state or local laws and regulations.
What we’re saying is, within the Tasmanian rules in your own planning code:
- To expand the clauses around what is minor communications infrastructure – that will streamline the rollout;
- To retain those environmental and heritage assessments but to make them much more targeted to areas of actual risk rather than generic;
- Again, to amend the code to bring in the idea of resilience, so multiple routes can be considered as part of one deployment rather than individually ‑ that would, again, speed up the process; and
- to recognise route diversity when you need to build multiple routes within the planning code, rather than having to do four separate applications to be able to do it wholesale.
All of these things, you’re keeping the protections in place, you’re just cutting through the red tape a lot more quickly.
Ms BADGER – It might be good if there are any very Tasmanian-specific examples in relation to that, because I think the view of the community is that they’re quite surprised how easy it is for telecommunication approvals to go through. That’s certainly what we saw with the Marinus Link cable. There were different parts of the process in relation to the electricity connection versus the data connection on-island, and certainly in St Helens there’s been concerns over the size of the tower and the proximity to the community. So, a specific Tasmanian example where the current planning has actually stopped large investment or large connectivity would be good.
Just the other aspect of this – because I notice we’re short on time, I don’t think we’ll get another round in – is in relation to the backup capacity required for large-scale AI factories, be it mobile, wireless or satellite backup that’s required. Can you talk us, if you can, through the practicalities of how that would work if we had a large-scale, say, AI factory or data centre in remote Tasmania, where there isn’t much, if any, connectivity currently.
Mr COLEMAN – For your first question, we will take that one on notice and we’ll come back with some specific Tasmania examples of where we’ve seen those roadblocks.
On the question of backup, when it comes to data centres, it’s really fibre or nothing. They are so high-capacity, they have such a massive appetite for very high-capacity connectivity, that the real backup is having multiple, geographically diverse fibre routes to ensure that data centre stays online all of the time.
Now, it’s different for consumers and businesses. For consumers and businesses that want to access AI applications, Low Earth Orbit satellites such as Starlink, Amazon – LEOs coming into market and others – they will provide fantastic connectivity for consumers that might have had really limited options in the past, if they only had mobile networks or traditional satellite networks. So, I think from a consumer perspective, certainly having multiple mobile networks and Starlink and other LEO satellites will really help solve that problem – but for data centres and providers themselves, it’s got to be fibre.
Ms BADGER – Thank you for that clarification.
Ms OGILVIE – Thank you very much. I would like to ask a question on digital economy. You mentioned specifically the uplink/downlink, upload/download question and, as I understand it, there is a big shift to people doing more on their phones and sending it off, a lot of it AI-driven. Do you have any particular oversight, or maybe you could even take it on notice, on what the economic development opportunity for digital and IP‑driven businesses in Tasmania ‑ we have screen, music, programming, university research, et cetera ‑ what that will look like, or could look like, if we get this right. Telco is one piece of it, the AI is one piece, but, you know, traditional research is obviously something we do well. I just thought as a peak organisation you might have some thought leadership pieces around that.
Mr COLEMAN – I will take that one on notice. It’s not something that we have developed ourselves, but I certainly know that other peak bodies have looked into those things. We can certainly come back to you for some information there that we haven’t looked at, how other sectors are looking to take it up, whether it’s the arts or small businesses or all those things. Other bodies, perhaps the Tech Council of Australia springs to mind, they have done a lot of good work in that regard. We have looked more specifically, as you would expect, at the use of bandwidth on telecoms networks. But we’ll take that on notice and see what we can come back to you with.
Ms BURNET – There’s significant community interest in AI data centres and the development applications that are being applied for across various parts of Tasmania. What do you think is peak AI data centre for Tasmania?
Mr COLEMAN – Well, how long is a piece of string? That’s a very tricky one to answer. We do see that there is a massive amount of investment appetite out there. Tasmania does have some unique opportunities. Data centres are hot. Cooling is a real factor when it comes to the costs involved in data centres. That’s what they need energy for. They are giant refrigerators, and so a more temperate climate is certainly attractive for data centre investment, but really, aside from energy and water for those cooling requirements, it is telco connectivity. So, I would say to your question, ‘what does peak data centre look like for Tasmania’, it is how quickly Tasmania can get more fibre in the ground and more fibre in the sea to get it back connected to other networks. That is a considerable constraint for investment in Tasmania where historically you haven’t had that inter-capital connectivity that other states have enjoyed.
Prof RAZAY – I am just wondering why we build so many data centres in cities scattered everywhere near the suburbs? If you have an excellent telecommunications network, why can’t we find an area outside of the city, build as large as you want and that will have less impact on the community, on amenity and on the environment?
Mr COLEMAN – Really relevant question that’s being asked in all of the jurisdictions looking at AI and data centres right now. While we often have very good telecommunications infrastructure in those capital cities, it dries up relatively quickly once you get outside those metropolitan areas. In the telco sector we refer to it as ‘connected buildings’. How many buildings in that area have competitive fibre going into the basement of that building that you can connect to? The more fibre you’ve got, the more connectivity that is available, the better the conditions to go and invest in that data centre.
Outside of capital cities, you typically have inter-capital routes. That is like the major superhighways that go from Sydney to Canberra to Melbourne. But only in a few places do those fibre networks break out. They have an off-ramp going off that superhighway. That is a real bottleneck for data centres being built in regional areas. It does require significant investment to build those off-ramps from those into capital fibre routes, but data centre investment could be a real driver to get more of those off-ramps from those into capital fibre routes. Ultimately, it just comes down to competition and availability of that connectivity. That’s a real driver for where you want to geographically locate a data centre.
Ms DOW – I just wondered if you’d made a submission to the Tasmanian government’s draft Statement of Expectations paper that closed recently?
Mr COLEMAN – No, we had not, but that’s something that perhaps we should look into if there’s an opportunity to do it later. But no, I wasn’t aware of that one.
Mr SHELTON – I know you touched on this a minute ago with Ms Badger’s question. Given this expansion that’s happening and fibre being the physical thing in the ground, satellite phones – and I have no doubt there will be investigations on how we can get around the need for this physical infrastructure versus some other physical infrastructure. Given change is always driven by need, is there a chance and what do you foresee in the future as far as how much of this is carried by in-ground infrastructure versus either mobile phone or, in fact, satellite phones?
Mr COLEMAN – It really comes down to the use case, that for the data centre itself it has to be fibre because it is such high-capacity processing and storage which is occurring in the data centre and so much of the communications is between one data centre to another data centre. If you’re going from a data centre in Sydney, Melbourne, Canberra to Hobart, it has to be via fibre.
However, for the consumers’ use, you are right. There’s going to be a diverse range of ways that consumers access AI, whether it’s an AI application like Siri or Claude or Copilot that’s on their mobile phone or a tablet computer that they’re doing on the go. Not only are we seeing an increase in AI traffic on mobile networks, but increasingly as we have services like Starlink and other LEO satellites that will be launched in coming years, yes, absolutely, we expect that those satellite services, particularly for people in regional and remote locations, there’s plenty of premises like that throughout Tasmania that will be using satellite mobile networks to access AI. As far as the data centres themselves go, they are very heavily dependent on fibre. I guess the answer is it’s both. It’s not one or the other.
CHAIR – I’ll just make the point that Tony Abbott’s comment in 2013 hasn’t aged well, about ’25 megabits per second being all that we need’. What jurisdictions globally are doing this best, Mr Coleman, in terms of the coordination between telecommunication infrastructure and data centres, and power and water. It’s happening globally, I know it’s happening fast, but who’s doing it best?
Mr COLEMAN – Sure, look, to be honest Chair, our focus has been on the Australian market rather than international. We’re wrestling with planning rules in-between our own states and our own local governments. We can certainly give some more information about which areas of Australia are doing it best, but we haven’t looked deeply into those international comparisons of how they’re reforming planning rules, I’m sorry to say.
Ms BADGER – I was just wondering if – it kind of builds on what the Chair has asked as well, and perhaps if you could take it on notice that would be helpful as well. It’s similar to what we asked the witness prior from the University of New South Wales about what other jurisdictions are also doing it well, so that in regional and remote areas – and they are having this fibre put in as new infrastructure, be it to facilitate a large AI factory or data centre – how it is then best that a percentage of that use is set aside to benefit that local community that it is then also connecting – which is obviously particularly important in a place like Tasmania, where we don’t have that that broad-scale rollout. If there are any other jurisdictions that are managing that balance well for the community, as well as the developer, that would be really helpful.
Mr COLEMAN – Yes, we can certainly take that one on notice. I would simply make the comment that getting that fibre in the ground is the best possible outcome in and of itself. When you’ve got that fibre built to go out to a data centre, the community automatically benefits because as soon as the infrastructure has been invested in, it’s usually the data centre operator or the tenant in that data centre – they’re footing the bill for it. The bill is not being paid for by the community or by the consumer. When that fibre is built, then you’ve got the opportunity to make the business case to break out that fibre, to make that off ramp into a local community. Maybe it supports a new mobile tower, maybe it supports a new LEO (Low Earth Orbit) satellite ground station in a remote area. Ultimately, making it as easy as possible to get that fibre built in the first place, ultimately it’s Tasmanian end users and communities that will benefit from that infrastructure.
CHAIR – All right, look, thank you both for your time. Thank you for your appearance. What you have said to us here today is protected by parliamentary privilege, and once you leave the table, you need to be aware privilege does not attach to the comments you may make to anyone, including the media, even if you are just repeating what you said to us. Do you understand that?
Mr COLEMAN – We understand, thank you.