Cyber Uncut

From AI-driven workloads and bank job cuts to attackers exploiting vulnerabilities at machine speed, David Hollingworth and Daniel Croft unpack a turbulent week in cyber security – including Australia's latest threat statistics and proposals to protect artists from AI-generated imitations. The Finance Sector Union says AI adoption is increasing workloads and stress, while job security falls amid thousands of bank cuts. Meanwhile, Arctic Wolf reports that AI autonomously resolves 60 per cent of security investigations, although human analysts remain involved in almost all cases – highlighting the importance of oversight as automation expands. PHP vulnerabilities have been added to the US Cybersecurity and Infrastructure Security Agency's Known Exploited Vulnerabilities catalogue, with AI helping attackers exploit longstanding flaws. The latest Australian Signals Directorate cyber threat report also points to rising data breach notifications, while four Australian organisations were hit by Qilin ransomware in just four days. There were positive developments, too. Commonwealth Bank of Australia and Deadly Coders are helping 5,000 First Nations students build AI and digital skills, while UNSW academics have proposed a Copyright Plus framework that would give artists greater control and potential remuneration when their work is used to train AI. Enjoy, The Cyber Uncut team

From AI-driven bank job cuts and online radicalisation to critical PHP vulnerabilities and ransomware gangs using AI to analyse stolen data, David Hollingworth and Daniel Croft unpack a week where artificial intelligence is reshaping both business and cyber crime. AI is transforming banking, but the benefits come with a human cost. On this week's Cyber Uncut, David Hollingworth and Daniel Croft examine Commonwealth Bank's rise to become Asia-Pacific's leading AI bank, alongside mounting job cuts across the sector. With banks struggling to guarantee job security and AI operating costs climbing, could some businesses soon find it cheaper to hire people than run the technology? The cyber threat landscape is getting uglier, too. Two critical remote code execution vulnerabilities in PHP have put potentially millions of websites at risk, with outdated systems particularly exposed and AI accelerating the development of working exploits. The pair also examine how ransomware gangs are using AI to analyse stolen files and build more targeted extortion campaigns, while warning that online networks and AI-generated propaganda are contributing to radicalisation and real-world harm. There are signs of progress amid the gloom. Westpac New Zealand is seeing growing adoption of open banking, giving customers more control over secure data sharing and payments. The hosts also explore the growing backlash against data centre developments, why community consultation matters, and Gartner's predictions for 2027 – including the decidedly cyberpunk prospect of "wetware-as-a-service", using biological systems for computing. Tune in for the stories shaping the intersection of AI, business, and cyber security. Enjoy, The Cyber Uncut team

From banks battling big tech for AI talent to ransomware attacks exposing sensitive medical records, David Hollingworth and Daniel Croft unpack another eventful week in cyber security – including why your supermarket fridge might be a bigger security risk than you think. AI is reshaping everything from banking to elections, but are organisations keeping pace with the risks? Hollingworth and Croft explore the race for AI talent, the collapse of Firmus' proposed $44 billion ASX listing, and growing concerns about AI-generated deepfakes in electoral advertising. With AI adoption racing ahead of governance, the pair examine the challenge of balancing innovation, economic opportunity and accountability. Then it's on to the week's cyber crime headlines, from French supermarkets hit by compromised refrigeration systems to ransomware groups claiming vast troves of sensitive healthcare data. The incidents affecting a pair of Aussie healthcare providers highlight the risks of third-party providers and compromised back-ups. Hollingworth and Croft also discuss why connected operational technology needs stronger security controls and why organisations must treat their suppliers as part of their own cyber perimeter. There's good news, too. Australia's new ASD Essentials framework marks an evolution of the Essential Eight, reflecting a threat landscape increasingly shaped by AI. And in a standout example of AI used for good, Indigenous rangers in Arnhem Land are using MegaDetector to streamline wildlife monitoring, cutting the images requiring analysis by 81.4 per cent. Tune in to Cyber Uncut for the latest developments, sharp analysis and a reminder that technology's biggest risks and most promising applications can emerge from the same place. Enjoy, The Cyber Uncut team

Cyber Uncut returns to discuss AI giant conspiracies, data centre impacts, and the dead internet theory. Are AI giants like OpenAI having their agents hack things on purpose? Senior journalist Daniel Croft and managing editor Liam Garman touch on the theory that these tech giants want their AI models hacking, as any publicity is good publicity. The two start by discussing dead internet theory and bots causing division, starting with the 2016 election. They then discuss local hacks and the Commonwealth Bank of Australia's AI copyright changes. Finally, the two talk about data centres and the impact they could have on markets. Are data centres set to cause a "mining boom" effect in small, regional towns? How are rental and housing markets going to be impacted? Should data centre developers be building housing right next to their projects? The answer might not be as clear as you might think. That's this week in the Cyber Uncut podcast. Enjoy, The Cyber Uncut Team

Information has always shaped power and influence – but in the age of artificial intelligence, the speed, scale and sophistication with which narratives can be created and deployed has changed dramatically. In this episode of the Contested Ground podcast, Phil Tarrant and Major General (Ret'd) Dr Marcus Thompson explore how the tradecraft of information warfare is increasingly intersecting with the commercial world, and why influence, narrative and information advantages are becoming critical strategic capabilities for modern businesses. They unpack the parallels – and important differences – between military information operations and commercial competition, from shaping perceptions and gaining decision superiority to defending against misinformation and responding to crises at speed. With AI now enabling highly targeted content to be created in hours rather than weeks, Tarrant and Thompson examine what this means for business leaders, why communications can no longer be left solely to marketing teams, and why organisations need the people, systems and strategy to compete in an information environment moving faster than ever before.

Cyber Uncut returns with an AI hacking incident at Medicare, ransomware drama, banking breaches, and the surprisingly serious security lessons from getting AI "drunk". Would you let an AI agent apply for a loan on your behalf? New research suggests 59 per cent of Australians would – but Cyber Daily's David Hollingworth and Banking Daily's Daniel Croft aren't convinced. The pair break down the growing willingness to hand AI sensitive financial data and decision-making authority, and why giving an agent autonomy creates a very different risk profile from simply asking ChatGPT to find you the best deal. The conversation then turns to the OpenAI agent that accessed the Australian government's Medicare statistics service, the fallout from the incident, and what happens when AI systems pursue goals with loose guardrails. From ShinyHunters' extraordinary clash with the FBI to Revolut's accidental disclosure of customer data, Hollingworth and Croft examine how AI-enabled threats and old-fashioned social engineering are converging – while ransomware groups continue to target organisations holding highly sensitive data. And then things get weird. Researchers at UNSW have demonstrated how prompting or fine-tuning AI to behave as though it were intoxicated can weaken its safeguards and make it more willing to overshare. Beneath the laughs is a serious warning: protections that work under normal conditions may not hold when an AI is pursuing a goal, adapting its behaviour, or encountering obstacles. Tune in for the latest Cyber Uncut conversation between David Hollingworth and Daniel Croft. Enjoy, The Cyber Uncut team

Artificial intelligence promises to transform productivity, business and government – but what happens when the technology starts making decisions its human operators never intended? In this episode of the Contested Ground podcast, Phil Tarrant is joined by Major General (Ret'd) Dr Marcus Thompson to examine the rapidly emerging security risks surrounding AI, from autonomous agents exploiting vulnerabilities to the growing challenge of keeping increasingly powerful systems inside appropriate guardrails. With businesses racing to deploy AI and governments grappling with how – or whether – to regulate it, Tarrant and Thompson explore whether Australia is moving fast enough to understand the threat, protect critical systems and ensure the technology designed to make us more productive doesn't become a new contested ground itself.

HPE's Matt Hanmer says organisations must break down the traditional divide between network and security teams as AI accelerates attacks and infrastructure becomes increasingly distributed. Organisations need to bring network and security operations together to keep pace with increasingly distributed infrastructure and faster-moving cyber threats, according to HPE APJ regional director Matt Hanmer. Speaking on the Cyber Uncut podcast, Hanmer said the traditional separation between teams is becoming unworkable as organisations connect more people, locations and applications while facing pressure to do more with less. Treating the network as a security sensor can help organisations detect threats while reducing duplicated tools, costs and operational complexity. AI is adding urgency to the shift, with attackers increasingly automating activity that once took months and compressing parts of the attack lifecycle into hours or minutes, Hanmer said. Defenders need to automate repetitive security operations while keeping humans responsible for oversight and intervention. Rather than keeping people "in the loop" for every decision, Hanmer advocates putting humans "on the loop" – setting rules and parameters for automated systems while ensuring security teams can intervene when necessary. Hanmer will discuss HPE's approach further at HPE Networking Day in Sydney on 7 October, where attendees can see its networking roadmap, demonstrations, and meet its global leadership team. Enjoy, The Cyber Uncut team

Australia and the United Kingdom share one of the world's deepest strategic relationships – but in an increasingly contested geopolitical environment, that partnership is taking on new significance. In this episode of Contested Ground, Phillip Tarrant and Major General (Ret'd) Dr Marcus Thompson are joined by acting British high commissioner to Australia Dr Brian Jones to unpack the future of AUKUS, the UK's growing Indo-Pacific focus, and the security challenges binding the two nations more closely together. They also explore the enormous industrial opportunity behind AUKUS, the role Australian businesses and technology can play in UK and allied defence supply chains, and why economic, technological, and national security cooperation is becoming increasingly inseparable. It's a wide-ranging discussion on deterrence, defence industry, sovereign capability, and the enduring trust underpinning the Australia–UK relationship. Enjoy the podcast, The Contested Ground team

Genetec's Lee Shelford says AI and cloud are giving security teams faster investigations while turning physical security systems into tools for broader business operations. Organisations should stop treating AI as a box-ticking exercise and focus on where it can deliver measurable operational gains, according to Genetec's Lee Shelford. Speaking on the Cyber Uncut podcast with David Hollingworth, Shelford says AI-powered video analytics can cut investigations from hours to minutes, allowing operators to find people, vehicles, objects, and behaviours using natural-language searches and respond while incidents are still unfolding. Shelford urges organisations to look beyond traditional surveillance, with Genetec combining video, access control, vehicle recognition, communications, and IoT data through a single interface. The technology can also support use cases beyond security, including bushfire detection, remote monitoring, car park management, and improving the movement of people and vehicles through airports and shopping centres. According to Shelford, remote management can eliminate the need for onsite upgrades and configuration, while AI can reduce the manual tuning previously required as conditions change. But cyber security must remain central, with Genetec having pursued Australian requirements, including IRAP assessment and Essential Eight alignment. Enjoy, The Cyber Uncut team

Phil Tarrant and Major General (Ret'd) Dr Marcus Thompson examine the growing debate over Australia's economic direction, drawing on Chris Uhlmann's recent article in The Australian on the "Argentinisation" of Australia. They unpack the risks posed by weak productivity, rising regulation, higher costs, and falling investment – and ask what sustained economic underperformance could mean for housing, business confidence, and the nation's broader strategic strength. Because if Australia enters an increasingly contested world from a position of economic weakness, its ability to fund defence, build sovereign capability, and protect its national interests becomes a much bigger question.

AI is pushing financial institutions to modernise, but the answer is not to replace everything. Banks should focus on the technology, data, and processes that directly support business outcomes. AI is putting modernisation back on the executive agenda as legacy systems increasingly limit banks' ability to adopt AI, access data in real time, and respond to regulatory and resilience demands. Atos executives Dee Lawler and Matthew Hughes argue modernisation is now a business issue, not simply a technology refresh. "The objective to modernise isn't to modernise the technology or the most technology," Hughes says. "It's to modernise the technology that matters most." But banks should resist the temptation to make AI the answer to everything. Leaders should first identify the business outcome they want, then determine what is holding them back – whether that is technology, data, processes, skills, or risk. Incremental modernisation can reduce operational risk while allowing organisations to measure returns, with resilience and business continuity built into the design. The takeaway for financial services leaders is simple: modernise where technology is constraining growth, efficiency, customer experience, or resilience – not simply because AI has created a sense of urgency. Enjoy, The Cyber Uncut team

Daniel Croft and David Hollingworth break down the essential AI and security news of the week, from sovereign capability to cyber crime, and fighting property scams. Hollingworth and Croft open the podcast with a breakdown of why Australian banks are standing up a new partnership program to promote the growth and adoption of local AI innovators – and explain why it's an important step in Australia's AI journey. At the same time, NSW has said it is taking steps to combat the rise of AI in classrooms by tackling the technology and its impact on the HSC. It's a good move. It's sadly been another bad week for Australian data finding its way onto the dark web via ransomware actors, with patient data once again compromised. But despite the hackers' claims that they hit NSW Health, the department has denied any such intrusion; the threat actor is definitely boasting, but the data is nonetheless real – Cyber Daily just doesn't know where it came from. And, finally, we sit down with PEXA's chief information security officer, Graham Fairley, to talk property scams: how they work, what a recent survey reveals, and how property buyers can protect themselves. Enjoy, The Cyber Uncut team

A dramatic political shift in Western Australia has put the major parties on notice and raised a much bigger question: what happens to Australia's national resilience when voters stop trusting the institutions that govern them? Hosts Phillip Tarrant and Major General (Ret'd) Dr Marcus Thompson examine the rise of populism in Australia and why growing political polarisation, cost-of-living pressures and economic frustration should be viewed through a national security lens. They unpack the erosion of trust in government and institutions, the weakening of Australia's traditional political centre and the risk that social division creates opportunities for foreign interference, information warfare and malign influence. Finally, the pair look north to the Pacific Islands Forum, currently underway, and Australia's increasingly contested strategic environment. With China seeking greater influence across the region, Taiwan remaining a flashpoint and several Pacific leaders absent from the forum, they discuss why Australia can no longer take its relationships in the Pacific for granted. Enjoy the podcast, The Contested Ground team

Join David Hollingworth and Daniel Croft for this week's round-up of cyber security and artificial intelligence news! The pair kick things off with the startling revelation that young Australians are turning to AI for their financial advice, as cost-of-living pressures bite. Can AI help get your finances in line? The jury is very definitely out. Then it's on to the impact of cyber crime on Australian businesses, with a relative newcomer ransomware groups targeting Aussie car and automotive dealers. Is it happenstance, or have the hackers found a vulnerable third party, in common with all the victims, to exploit? Finally, Hollingworth and Croft take a look at massive changes coming to Facebook and Instagram as Meta settles a US court case alleging the company's actions to create addictive content to capture the teen market. A definite good thing, but Meta wants the industry to go further. Enjoy, The Cyber Uncut team

Microsoft 365 is the backbone of many Australian businesses – but are organisations doing enough to secure the platform itself? Tune in to David Hollingworth and CoreView's vice president of APAC sales, Andrew McAllister, as they tackle the risks and what businesses can do to build genuine resilience. Microsoft 365 may be ubiquitous, but that doesn't mean it is secure by default. Hollingworth and McAllister explore the often-overlooked risks lurking in tenant configurations, administrative privileges and guest accounts, and explain why organisations need to take greater responsibility for how their Microsoft environments are configured and governed. The conversation also examines how attackers are exploiting compromised admin accounts to manipulate settings, bypass security controls, and gain deeper access to sensitive data. McAllister explains why the traditional Microsoft 365 admin model can create an unnecessarily large blast radius, and how organisations can apply least privilege, role-based access and tighter identity controls to reduce their exposure. Finally, Hollingworth and McAllister examine what a mature Microsoft 365 resilience strategy should look like – from establishing a secure configuration baseline and detecting configuration drift to automated remediation and recovering the tenant itself after a serious compromise. Listen to the full episode to find out why backing up your data isn't enough – and what organisations need to do to protect the "glass" holding it. Enjoy, The Cyber Uncut team

Malicious browser extensions, AI-powered phishing, and the evolution of Australia's Essential Eight are all putting renewed focus on proactive cyber defence. Hear ThreatLocker's lead cyber security engineer, Kieran Human, explain why organisations need to get the basics right before attackers get the chance to exploit them. Malicious browser extensions are giving attackers another route into corporate environments, with phishing campaigns tricking users into installing tools capable of running arbitrary code. Cyber Daily's David Hollingworth and Human break down how the so-called Edgecution attack method works, how initial access brokers fit into the ransomware ecosystem, and why even highly technical employees can fall for increasingly convincing social engineering. The conversation then turns to what defenders can actually do about it. Human argues that security teams need to move beyond relying solely on detection and response, with application allow-listing and other proactive controls providing a critical layer of protection when users inevitably make mistakes. Finally, Hollingworth and Human examine the future of Australia's Essential Eight, including the move towards the new Essentials framework and potential guidance for enterprise, cloud, and AI environments. Enjoy, The Cyber Uncut team

In this episode of Cyber Uncut, Cyber Daily's Liam Garman is joined by Alex Nehmy, who explains why securing critical infrastructure requires a different approach to traditional IT security. Discover why operational technology has become one of the highest-value targets for sophisticated threat actors, how attacks on OT differ from conventional data breaches, and why visibility into industrial environments is the foundation of effective defence. The discussion explores real-world examples of cyber attacks causing physical disruption, the growing risks posed by supply chains and connected devices, and why the idea of a truly "air-gapped" network is often more myth than reality. The conversation also examines how AI is reshaping both offensive and defensive cyber operations, the challenges organisations face in protecting modern extended OT (xOT) environments, and the practical steps security leaders can take to improve resilience against nation-state threats. Whether you're responsible for IT, OT, or critical infrastructure, this episode offers valuable insights into one of cyber security's most important and rapidly evolving battlegrounds. Listen now to hear how security leaders can strengthen their OT defences, reduce risk, and stay ahead of increasingly sophisticated cyber adversaries. Enjoy, The Cyber Uncut team

Dr Ross Babbage warns that without urgent economic reform and industrial recalibration, Australia risks becoming a client state of Beijing. Returning to the Defence Connect Podcast, strategic analyst Dr Ross Babbage delivers an unsparing diagnosis: Australia is drifting towards a future of strategic weakness, economic vulnerability and de facto dependence on Beijing – unless the nation urgently reverses course on industrial policy, governance efficiency and the information asymmetry crippling public debate. In a wide-ranging conversation with host Steve Kuper, Babbage articulates an emerging geopolitical fracture: the world is crystallising into distinct economic camps. The first, led by the United States and including Western democracies, is erecting selective tariffs and deepened trade ties. Camp two is China and its authoritarian satellites. A third comprises developing nations watching carefully before committing. Australia's position, Babbage argues, remains fatefully ambiguous. While virtually every Western ally has taken protective measures against Chinese manufacturing dumping and industrial subsidy – a regime he quantifies at eight to 18 times the support other nations provide – Australia has remained conspicuously reticent. China's leverage remains decisive: our iron ore and coal exports are non-negotiable parts of Beijing's strategic calculus. The deeper problem, he stresses, runs through economic governance itself. Australia has deindustrialised deliberately over three decades, shed critical capabilities in steel and advanced manufacturing, surrendered energy cost advantage and starved investment in domestic capability. Without deliberate recalibration – smaller government, redirection of trade partnerships, strategic investment incentives – Australia defaults towards clientelism. Equally critical is the communication void. Defence industry leaders, board chairs of major firms and even policymakers routinely tell Babbage they were unaware of the scale of Chinese subversion, intellectual property theft and coordinated disintegration operations aimed at fracturing Western cohesion. The path forward, he argues, demands frank public briefings, ruthless priority setting and abandonment of exhausted economic orthodoxies. Without it, Australia becomes the strategic backwater Beijing expects. Enjoy the podcast, The Contested Ground team

In this week's podcast, Cyber Daily's Daniel Croft and David Hollingworth argue for sovereign Australian AI, catch up with the Origin Energy data breach, uncover a Court Services Victoria hack, and ponder the efficacy of taking a foreign chat platform to court over terror content. Should Aussie banks have access to Australian Taxation Office data to stop fraud? That's how the podcast kicks off, alongside the revelation that fraudsters are using AI to fake documents to secure massive bank loans. For Hollingworth and Croft, the answer is a resounding yes, which is also the case when it comes to sovereign AI capability, as Australian company Maincode brings its own coding agent online. This week, Origin Energy revealed the extent of its recent data breach, while Court Services Victoria disclosed it had fallen victim to prolific hacker 2019. It's been a torrid time for data security in Australia, with even an insider threat rearing its head in a data breach at NSW Health. Finally, the pair discuss whether or not eSafety can effectively sue chat platform Telegram over pro-terror content, while the Australian Communications and Media Authority said it is suing Optus over its Triple Zero failures last year. Just another week in cyber security. Enjoy, The Cyber Uncut team.