This product has already been purchased.
Physical AI and quantum supercomputing are transforming our lives. We unpack five tech frontiers every board should have on its radar.
1. Robotics – physical AI
Australia was the first nation to automate our ports and mine sites, we lead the world in field robotics innovation, and yet we rank near the bottom of 35 countries measured by robot adoption per worker, according to International Federation of Robotics statistics. For company directors, this gap represents not a warning, but an opportunity.
For decades, robotics meant enormous capital outlays. Autonomous load haulage vehicles at mine sites and associated infrastructure can cost upwards of $100m per mine site. However, the surge of global investment into humanoid and service robots is driving the price of smaller robots to a point where meaningful experimentation is now accessible without bet-the-company risk.
“Directors who dismissed robotics five years ago should look again. The price of hardware has changed profoundly,” says Dr Sue Keay, director of the UNSW AI Institute and director and founder of the Robotics Australia Group, the peak body for the robotics industry.
The sweet spot for early adoption is clear. Take the physical, repetitive or injury-prone tasks away from humans and reallocate that human capacity to higher-value work. But this is not straight personnel replacement.
“Successful implementation requires genuine understanding of your business processes and a willingness to redesign them,” says Keay. “Directors must ask, do we know our workflows well enough to deploy automation intelligently?”
Deloitte’s recent State of AI in the Enterprise report found that 57 per cent of Australian companies are already employing physical AI to some extent, with adoption projected to exceed 80 per cent within the next two years.
Physical AI in action
SwarmFarm Robotics is a Queensland-born agtech company. Their flagship SwarmBot is an advanced autonomous farming robot that independently navigates fields using GPS with computer vision-based weed detection to perform tasks like seeding, spraying and weed control. The technology is reported to deliver a reduction in chemical usage of up to 80 per cent through precise application, while using lighter machinery reduces soil compaction and improves crop yields. Their open developer platform, SwarmConnect, acts like an app store, enabling innovators worldwide to build new agricultural technology compatible with SwarmBots.
Advanced Navigation was founded by engineers Xavier Orr and Chris Shaw, now CEO. The Sydney-based technology company makes robots, drones, vehicles and machines capable of operating alone in difficult places. Its underwater robot can swim on its own, take high-quality video and create 3D maps to study seabeds or coral reefs.
Their autonomous vehicles can enter contaminated sites or search for any signs of life after underground incidents, mapping environments or detecting hazards. In March, it raised US$110m in a series C funding round.
Co-founder Orr recently set up Anitron, developing humanoid robots to solve labour shortages in construction, aged care and manufacturing. Anitron’s foundation model is based on insect intelligence-derived AI (insect neural mechanisms are translated into AI algorithms and robotic systems) which Anitron claims requires substantially less processing power than alternatives. It runs directly on the robot rather than in a data centre. Beta trials began this year.
“Abi” (empathetic robot companion in aged care) was developed by Andromeda Robotics in 2022. Since then, Abi has logged more than 3500 hours of companionship across elder care facilities in Australia, conversing with residents in care facilities in 90 languages. Abi remembers previous conversations and, in group settings, can lead tai chi sessions, run trivia competitions and play music. Founder and CEO Grace Brown leads the team working across Australia and is now expanding into the US.
FBR specialises in designing technology for the construction, maritime and industrial sectors. Their Hadrian X robotic bricklayer uses a 3D model of a house to place blocks exactly where they need to go, applying a special adhesive instead of traditional wet mortar. It can lay around 300-360 blocks per hour, which is what a person can on average lay in one day. And Hadrian X can work in any weather conditions, 24/7.
The welding arm, Mantis, uses lasers to “see” where it needs to weld and can handle complex jobs over a large area, remaining stable across its eight-metre reach.
Aiming to commercialise globally, FBR recently appointed former Brickworks CEO Lindsay Partridge AM FAICD as an independent non-executive director.
What boards get wrong
The most common mistake directors make is demanding an 18-month return on investment. For a meaningful robotics implementation, that timeline is unrealistic and will kill good projects before they mature, says Keay.
“Boards should instead frame robotics the way they frame any infrastructure investment – with a longer horizon and a focus on capability building, risk reduction and competitive positioning.”
2. AI-enabled cyberthreat detection
The cyberthreat landscape has fundamentally shifted, creating urgency for AI-powered threat detection. The average time for an attacker to move from an employee’s laptop to infiltrate the company’s network or data bases dropped to just 29 minutes in 2025, a 65 per cent decrease on the previous year. The fastest recorded eCrime breakout time was just 27 seconds, according to CrowdStrike. Attackers are now using AI at every stage – reconnaissance, credential theft, evasion – and they are able to operate without compliance cycles, regulatory constraints or governance frameworks slowing them down.
The first lesson is that identity is the new perimeter, says Fabio Fratucello, Worldwide Field CTO for CrowdStrike. Attackers increasingly don’t “break in”. They log in using stolen credentials. Valid credentials generate less noise, making threats harder to detect. This is happening inside companies now, not in some future threat scenario.
The second is about complexity. Many organisations run separate tools for endpoint (physical device that connects to a network system), identity, cloud security and vulnerability management, he says. When these tools don’t talk to each other, defenders can’t connect the dots to detect cross-domain attacks. AI-driven platforms that unify telemetry across domains give security teams the ability to act at machine speed.
The third lesson concerns AI adoption itself, says Fratucello. Every company is already using AI, which expands the attack surface. Increasingly treated like digital employees, with access to sensitive systems and data, AI and AI agents can be manipulated through “prompt injection” where adversaries inject hidden instructions into GenAI tools to weaponise the very systems transforming how work gets done.
Don’t wait for the perfect solution
“Boards that wait for the perfect framework – or wait for the government to tell them what to do – will be significantly behind,” says Fratucello. “AI is here, so start to build visibility and security control capabilities with what’s available today. You might not have the perfect solution, but you’ll be on the front foot, as much as you possibly can be, if you have a solution that secures AI.”
3. AI healthcare diagnostics
AI is being used to provide speed and clarity in diagnostics throughout the Healius pathology network, says its chief operating officer, Puneet Nagi.
Healius is partnering with an international histopathology AI provider to scan tissue samples and overlay clinical information on images to support faster, more accurate diagnosis. Histopathologists review AI suggestions and can approve, reject or use them to initiate further examination.
Generative AI algorithms are used to scan high-resolution digital images of biopsy slides to detect subtle indicators of cancer that might be missed by the human eye. AI imaging tools can enhance an image to identify cancer-related biomarkers in tissue samples, increasing the chances of early detection. It’s also contributing to reducing the time required for diagnosis and allowing pathologists to focus on complex cases.
Agentic AI is not a magic wand, says Nagi. “We create the guardrails and scope, keeping humans in the loop at all times. We need to always know what the AI is doing, we don’t want to be surprised later.”
Nagi makes the point that boards need to understand the potential use cases of new technology, but also decide on the strategy of how it will be created and used. Boards should ask if companies need to build their own technology or partner with tech experts.
“We could write that software ourselves,” says Nagi. “We could build it, but the strategic choice we’ve taken is to use the best technology in the world, let the technologists create the best algorithms while focusing our energies on adopting these modern insights into our clinical pathways. We believe the value to the business is not so much in creating the new version of the algorithm, but in deploying that algorithm effectively so our people can work faster and better.”
“The more interesting part for me is the challenge that boards can create for executive teams,” he says. “If you’re presented with a technology transformation plan that will cost $50m and take 18 months to do, I’d ask, ‘How can you take this plan and turn it into a two or three-month outcome and just spend $5m?’ That wasn’t possible six or nine months ago. The scale of things you can achieve, and the speed at which you can achieve them, has absolutely escalated.”
4. Supercomputing
At the heart of scientific discovery, industrial modelling and, increasingly, AI development, sits supercomputing. Training large AI models, running simulations that underpin engineering and pharmaceutical research, processing the flood of data from radio telescopes or genomic sequencing are all supercomputing workloads.
The Pawsey Supercomputing Research Centre in Perth houses Setonix, one of the most powerful supercomputers in the Southern Hemisphere. It can analyse vast amounts of data in seconds, rather than years, to map the universe, or create detailed virtual models for medicine, energy and climate studies, such as predicting climate change impacts or designing new drugs.
Pawsey is a pioneer in combining quantum computing (QC) and AI technologies, having installed the world's first room-temperature, diamond-based quantum computer to work alongside Setonix. It is currently being used for drug discovery. It can enhance medical images for higher-resolution scans for earlier diagnosis of conditions including early-stage brain cancer. Future applications might include optimising renewable energy by better predicting energy production patterns. Used together, QC and AI can be applied to solve massive supply chain jigsaw puzzles in logistics, mining and space.
5. The AI–quantum convergence
However, with all eyes fixed on AI, opportunities arising from the convergence of QC and AI, could be missed. Directors need to start delving deeper says Pawsey CEO Mark Stickells AM FAICD, who is also a non-executive director of Science and Technology Australia (STA) and the Australian Research Council. He suggests injecting tech voices into board conversations. “Explore national infrastructure access programs, such as those offered by the Pawsey centre, which can connect commercial organisations to world-class capability.”
He also advocates supporting initiatives to build the pipeline of technically fluent governance leaders, including STA’s new Science meets the Economy program. “Treat R&D and STEM sector engagement as a strategic investment, not a cost,” he says. “The government’s recent Ambitious Australia report argues our country needs to be thinking about how to organise itself around innovation and the relationship between the sector and industry. It’s rich ground for us to have good conversations across boardrooms. To see the opportunities in that area, we may need different voices in the boardroom, but it also needs our science and technology sector to understand how business works.”
Organisations that are prepared will gain early advantage, and those that aren't will face significant cybersecurity risks, says quantum expert Dr Anthony Lowe GAICD.
“Unfortunately, the arrival of quantum computers will also mean most current encryption methods we’ve relied on for decades to safeguard online transactions and trade secrets will be compromised,” he says.
“The situation is particularly urgent for organisations such as banks and healthcare providers, whose data is likely to remain sensitive up to seven or more years into the future. Hackers are already copying data in transit and storing it until the advent of QC. Fortunately, post-quantum encryption algorithms have already been developed, which neither conventional nor quantum computers can break efficiently.”
Classical AI and machine learning thrive on large, structured datasets. But decisions in business and society involve sparse, complex, hard-to-standardise data. Health information can present unique challenges when there are minimal details to train a model on. Early results are limited but encouraging with regards to accuracy and reliability of predictions, says Pawsey, but more research is being done.
“Every traditional risk domain, financial, operational, regulatory, reputational, now has a compute and AI dimension underneath it, which makes it a platform, not a peer category,” says Stickells. He adds the critical difference is that cyber governance is often defensive. “Boards that stay illiterate will pay twice – in downside exposure and missed competitive upside.”
Whole-of-board literacy
The AICD’s resources and research on cyber risk has demonstrated boards need whole-of-board tech literacy.
Big tech isn’t underestimating the risk. Google has recently announced a significant acceleration of the incorporation of post-quantum cryptography (PQC) into its Android platform by 2029. This may well suggest that it believes commercially available QC is closer than many imagine.
“Company directors should be ensuring management assesses the organisation’s data time sensitivity and begins the transition to PQC now,” says Lowe.
This article first appeared as 'The next big tech' in the June/July 2026 Issue of Company Director Magazine.
Latest news
Already a member?
Login to view this content