OAIC Releases New Data: Cyber Attacks and Human Error Drive High Data Breach Levels

The Office of the Australian Information Commissioner (OAIC) has released its latest Notifiable Data Breach (NDB) statistics, covering the January – June 2025 reporting period. These figures underscore the ongoing threats facing Australian privacy, including cyber risks, human error, and vulnefqrabilities arising from outsourcing personal information handling.

Under the Privacy Act 1988 (Cth), the NBD scheme requires covered organisations and agencies to notify both the affected individuals and the OAIC if a data breach is likely to result in serious harm. These notifications must also include recommendations to help affected individuals respond to the breach.

January – June 2025 Snapshot

The latest report reveals that cyber risk remains a pressing concern for Australian organisations. Despite a 10% drop in overall notifications (532 breaches), the scale and impact of incidents remain significant. Malicious or criminal attacks accounted for the majority (59%), underscoring the sophistication of threat actors, while breaches caused by human error surged to 37%, highlighting persistent gaps in training and governance. On average, each cyber incident affected over 10,000 individuals, with the cost of a breach estimated at $4.26 million. Sectoral trends show health (18%), finance (14%), and government agencies (13%) as the most impacted, reinforcing the need for robust security controls and proactive risk management across critical industries.

Managing Third-Party Risks

The OAIC also highlighted the continuing challenges posed by data breaches involving third-party service providers. Organisations are responsible for the actions of these providers when outsourcing their personal information handling. This responsibility becomes even more critical as supply chains increasingly integrate advanced technologies, including artificial intelligence (AI).

Steps organisations could implement to reduce the impact of a supply chain risk include:

• engaging suppliers that have demonstrated robust security controls (e.g., ISO 27001, NIST CSF);
• incorporating contractual clauses addressing data collection, use, retention, destruction, and breach notification obligations. Contracts should also require adherence to AI governance principles where AI tools are deployed;
• ensuring effective oversight of third-party providers, which involves regularly conducting cyber security assessments and audits; and
• introducing AI-specific risk controls requiring suppliers to disclose AI usage in data processing and decision-making, and implement monitoring for adversarial attacks on AI models, which can compromise data integrity or expose sensitive information.

By embedding these measures into procurement and governance processes, organisations can strengthen resilience against both traditional cyber threats and emerging AI-related risks.

Since 30 May 2025, it been mandatory for organisations with an annual turnover of more than AUD $3 million to report to the Australian Signals Directorate (ASD) within 72 hours of making a ransomware or cyber extortion payment.

In December 2025, the Home Affairs Minister Tony Burke made a statement that during this period (May to December 2025) 66 Australian businesses had come forward to report making ransomware payments, however many businesses are still not reporting. The Australian National Cyber Security Coordinator Michelle McGuinness also stated that while paying ransom is not illegal to accommodate for “life or death scenarios” for critical infrastructure, paying ransoms feeds the cycle of criminality.

From 1 January 2026 the Department of Home Affairs will transition to Phase 2 of the ransomware payment reporting rollout, adopting a compliance and enforcement approach. It is expected that reporting business entities will be aware of and have integrated ransomware payment reporting obligations into their cyber incident response plans.

Recent warnings from ASIO Director-General Mike Burgess have attracted considerable attention, underscoring the growing urgency around Australia’s cyber defence posture. The Director-General cautioned that state-sponsored cyberattacks are becoming “devastating and disruptive” with threat actors probing Australia’s critical infrastructure including financial and telecommunications networks.

He stresses that such penetrations could allow adversaries to disable power grids, disrupt financial systems, or contaminate water supplies during moments of strategic tension, with some incidents carrying an estimated economic impact of up to $1 billion per breach (Ibid.). These comments reflect a recent attack detected in the UK where a state-sponsored threat actor used AI’s agentic capabilities to target technology companies, financial institutions, chemical manufacturers and government agencies.

This escalation is not limited to critical systems, Australia’s private sector is now firmly in the crosshairs of state-sponsored threat actors. The Director-General warned that foreign intelligence services are aggressively targeting companies to steal trade secrets, intellectual property, and sensitive customer data, often to give their domestic industries a competitive edge. Recent cases include the compromise of a major exporter’s network to gain leverage in contract negotiations and the theft of proprietary designs that enabled mass production of knock-offs, nearly bankrupting the original innovator. For business leaders, this means cyber security is no longer a technical issue, but a strategic priority tied to competitiveness, reputation, and national resilience.

While primarily aimed at raising awareness, the Director-General’s commentary aligns with broader policy discussions on strengthening the nation’s cyber security protections and the alliance between government and industry. With the mandatory ransomware and cyber extortion reporting in effect from 30 May 2025, such remarks intentional or not reinforce expectations that cyber resilience will remain a top-tier national priority.

Pro‑Russia hacktivist groups are increasing their focus on critical infrastructure worldwide, prompting coordinated warnings from international cyber authorities. A joint advisory issued in mid‑December by the Cybersecurity and Infrastructure Security Agency (CISA) and multiple global partners outlines a pattern of low‑sophistication but disruptive attacks targeting operational technology (OT) environments across a range of critical infrastructure.

The groups highlighted include the Cyber Army of Russia Reborn, NoName057(16), Sector16 and Z Pentest. Authorities have observed these actors scanning the internet for exposed OT systems and exploiting weak or default credentials to access devices used in industrial operations. Although they lack the capabilities of more advanced cyber actors, their activity has still caused disruptions that require operator intervention.

A recurring theme in the incidents is the misuse of publicly accessible Virtual Network Computing services. These services provide remote access to human machine interfaces used for monitoring and control. In many cases, they were exposed without adequate authentication or network segmentation. As a result, attackers were able to view or alter configurations and, in some instances, interrupt normal operations. The advisory notes that some incidents caused service interruptions and that a smaller number led to limited physical impacts.

Affected sectors include water and wastewater systems, food and agriculture operations and energy infrastructure. These environments depend on stable configuration and continuous system visibility. Even short interruptions can delay processes, require manual workarounds or temporarily reduce situational awareness. While some hacktivist claims appear exaggerated, international agencies have confirmed several cases involving real interference with OT devices.

Authorities emphasise that the scale of the issue stems from the widespread availability of internet‑connected OT systems and the simplicity of discovering them. Actors commonly scan for accessible devices, attempt logins with weak or default credentials and then publicise their claimed successes online to increase visibility.

The mid‑December advisory reflects a continued rise in opportunistic incidents involving pro‑Russia hacktivists. Although the methods remain simple, the impacts on affected systems can be significant for operators responsible for maintaining safe and reliable service delivery. The advisory reinforces that exposure of OT systems to the public internet remains central to these incidents and continues to be a priority for coordinated international monitoring.

On 20 November 2025, the Australian Government in coordination with the United States Department of the Treasury’s Office of Foreign Assets Control (OFAC) and the United Kingdom’s Foreign Commonwealth and Development Office, imposed sanctions on Russian cybercrime service providers. The measures target Media Land LLC, its sister company ML Cloud and two senior staff members Aleksandr Alexandrovich Volosovik (General Director of Media Land LLC) and Kirill Andreevich Zatolokin (runs ML Cloud).

Media Land operates as a ‘Bulletproof’ hosting (BPH) service, providing ransomware infrastructure to cybercriminals for attacks on critical infrastructure in Australia and globally. BPH service providers offer a digital platforms designed to evade detection and law enforcement takedown, enabling malicious cyber activity. The Australian Signals Directorate (ASD) provides more information on ‘Bulletproof’ hosting providers here.

Under these sanctions, it is now a criminal offence to engage in transactions or activities with the sanctioned entities or individuals, including using or dealing with their assets or providing assets. Penalties include up to 10 years imprisonment and/or heavy fines. The sanctioned individuals are also prohibited from entering Australia (Ibid.).

“These sanctions don’t just impose costs on criminals, they dismantle the infrastructure that enables cybercrime. By disrupting these networks, we make it harder for others to launch attacks and it strengthens Australia’s resilience against future threats.”

Deputy Prime Minister Richard Marles

This action marks the fifth time Australia has activated its autonomous cyber sanctions framework. It underscores a coordinated international effort to dismantling entire cybercrime networks rather than targeting individual hackers alone, serving as a proactive deterrent against future attacks.

Introduction

2025 marked a clear shift in cyber‑criminal activity, with attackers increasingly turning their attention from direct intrusions on individual organisations and instead targeting the technology supply chain.

When threat actors compromise trusted software vendors, service providers or other upstream dependencies, they can exploit that position to infiltrate downstream environments or access data entrusted to those providers. In such circumstances, losses can ripple through entire sectors, transforming a single upstream compromise into a large‑scale, multi‑party incident.

Here we look at how supply-chain attacks manifests in the technology sector, the nature and extent of the risks they create for organisations, and the key considerations for developing an effective mitigation strategy.

What Is a Supply‑Chain Attack in the Technology Sector

In short, a supply-chain attack occurs where a threat actor targets a third‑party vendor, service provider or technology dependency to compromise the organisations that rely on them.

Rather than attacking each organisation directly, the attacker exploits the implicit trust placed in upstream providers by leveraging software updates, managed services, integration points or privileged access arrangements to move into downstream environments.

Supply chain attacks are therefore uniquely positioned to impact multiple organisations simultaneously, amplifying both the scale and complexity of resulting incidents.

Core Risks Posed to Organisations

Several notable 2025 cyber incidents highlight the core risks that supply chain attacks create for organisations and providers.

Privacy and Regulatory Risk

Supply chain attacks can expose large volumes of personal, operational or confidential data, even when the affected organisation has limited direct control over the breach.

For instance, in October 2025, Discord experienced a data breach after attackers compromised a third-party support vendor. The attackers accessed user data, including personal details, limited billing information and some government-issued IDs, impacting millions of users.

Despite not being the source of the vulnerability, Discord was required to notify regulators and affected users, faced possible regulatory investigations in multiple jurisdictions, and attracted scrutiny from class action firms.

Operational Disruption and Business Interruption

Because supply chain attacks often originate upstream, they can introduce malicious changes directly into an organisation’s systems, bypassing conventional security controls. This frequently results in downtime, service outages and, in severe cases, sector-wide compromises.

For example, in mid-late 2025, attackers inserted harmful code into widely used NPM packages. Node Package Manager (NPM) is a system used to distribute and manage small pieces of software code, “packages”, for applications built with JavaScript. In plain terms, it’s a standardised distribution and filing system that lets developers reuse, update, and track third-party code instead of writing everything from scratch)

As a result, malicious updates automatically reached thousands of organisations during routine software updates, exposing them to data theft and unauthorised access.

The incident forced many affected businesses to roll back releases, refresh credentials, and temporarily shut down services for investigation and containment, highlighting the far-reaching operational disruptions and risks tied to third-party code dependencies.

Contractual, Commercial and Liability Exposure

When a third-party provider is compromised, organisations often encounter disputes over responsibility, indemnities, and the extent of a vendor’s contractual obligations.

Many suppliers attempt to limit liability to minimal amounts or exclude cyber risks altogether, forcing downstream organisations to bear substantial losses. This risk is illustrated by the August 2025 Salesloft Drift platform breach. Salesloft Drift, which connects with CRM and communications platforms like Salesforce, was compromised when threat actors gained access to authentication tokens, enabling unauthorised entry to customer environments and the export of sensitive data, including personal information and system credentials.

Although the breach originated at a third-party vendor, Salesforce subsequently faced claims from affected customers, including major US insurers and credit bureaus, underscoring the complex liability issues that arise following supply-chain attacks.

Mitigation Strategies for Organisations

The breadth of risks illustrated by recent supply-chain incidents makes clear that vendor‑originated breaches cannot be treated as outliers or hypotheticals. Organisations should adopt a deliberate and structured approach to managing third‑party exposure, that integrates governance, contractual protection and technical assurance.

Depending on an organisation’s size, structure and risk appetite, practical steps may include:

Strengthen Vendor Governance

Conduct structured due‑diligence on critical vendors, including assessment of security posture, access levels and dependency chains. Require evidence of controls rather than relying on assurances and mandate core security controls such as endpoint detection, patching, logging and monitoring and regular penetration testing.

Review Contractual Protections

Ensure contracts include meaningful (and enforceable) indemnities, mandatory incident notification timeframes, security obligations, liability caps tied to risk exposure, and rights to audit or assess the vendor’s controls.

Insurance Coverage Reviews

Assess how cyber, business interruption, crime and professional indemnities policies would respond to a vendor‑originated incident, including definitions, sub-limits, waiting periods and aggregation exposure.

Conduct Regular Tabletop Exercises

Simulate and run supply-chain breach scenarios with executives, legal, risk and technical teams to clarify roles, decision points and insurer‑notification obligations. Model the impact of a major vendor outage or upstream compromise to understand operational dependencies, recovery time objectives and alternative service pathways.

Maintain Visibility Over Third‑Party Dependencies

Catalogue all vendors and sub‑vendors (including open‑source components) to understand where hidden dependencies exist and how they connect into operational systems.

Conclusion

As organisations depend more on a growing network of software providers and cloud platforms, their security increasingly ties to third‑party risks. Resilience now requires visibility, strong contracts, and readiness to respond to upstream compromises.

Though supply‑chain risk cannot be eliminated, early investment in governance and preparation helps minimise disruption, meet regulations, and safeguard customer trust.

At WK, we support clients by reviewing vendor relationships, improving contractual protections, assessing insurance, and strengthening incident‑response plans. If you want to enhance your third‑party risk management, our team can help.

In October 2025 the Privacy Commissioner found that Vinomofo Pty Ltd (Vinomofo) breached the Privacy Act 1988 (Cth) by failing to take reasonable steps to protect personal information, contrary to Australian Privacy Principle (APP) 11.1. This decision followed a data breach affecting nearly one million individuals. The Privacy Commissioner directed Vinomofo to cease the practices that led to the breach and implement remedial measures within 90 days (Australian Information Commissioner v Vinomofo Pty Ltd [2025] AICmr 175.)

The breach

In September 2022, a threat actor accessed and exfiltrated a misconfigured cloud database used for data migration, exposing approximately 17GB of data, as reported in the determination issued by the OAIC. The compromised dataset included names, contact details, dates of birth, gender, addresses, and financial records for customers and members. The database lacked basic security controls: no isolation from the internet, no web application firewall, and no encryption enabled. The threat actor demanded a ransom, then posted the stolen data for sale on the dark web on 16 October 2022. A sample was subsequently sold on 20 October 2022.

The failures

The Privacy Commissioner determined that Vinomofo interfered with individuals’ privacy by failing to take reasonable steps to protect customer information. The investigation highlighted three major deficiencies:

• Inadequate technical measures: Critical gaps in security logging, lack of encryption for sensitive data, and poor cloud configuration during a migration project.
• Weak privacy culture and governance: Privacy was internally dismissed as “the boring stuff,” with outdated policies, inconsistent training, and a culture that did not prioritise accountability or risk awareness.
• Failure to act on known risks: Security shortcomings were identified at least two years prior to the breach, yet no remedial action was taken. Risks linked to cloud migration were ignored, and industry standards such as NIST CSF and ISO 27000 were not adopted.

The determination and remedies

The Privacy Commissioner concluded that Vinomofo’s steps were not reasonable given the:

• nature and volume of personal information held (928,760 records);
• company’s resources ($72 million annual revenue and 120 employees); and
• potential harm to individuals, including identity theft and dark web exposure.

Notably, whilst no penalty was imposed, Vinomofo was ordered to:

• implement stronger technical controls: enable security logging across AWS, enforce database access controls, and deploy monitoring for unauthorised activity;
• establish policies and procedures aligned with minimum industry security standards;
• engage an independent reviewer to assess cyber security capabilities and address gaps;
• submit the reviewer’s report and implementation steps within 14 days; and
• maintain ongoing compliance with APP 11 obligations.

Key takeaways

• APP 11.1 requires entities to take reasonable steps to protect personal information from unauthorised access, loss, or disclosure. Reasonable steps include technical and organisational measures, robust governance, and cultural prioritisation of privacy.
• Cloud migration projects pose heightened risks; entities must ensure proper configuration, isolation, and encryption of temporary environments.
• OAIC is increasingly focused on systemic privacy failures and organisational culture, not just technical controls.

Southern Phone, an Australian telecommunications company, received a record penalty of $2.5 million after the Australian Communications and Media Authority (ACMA) discovered 168 violations of anti-scam regulations between July 2024 and February 2025. These breaches allowed scammers to bypass identity checks, take over customers’ mobile numbers, and gain access to their bank accounts, leading to reported customer losses of at least $393,000.

In a statement released on 3 December 2025, ACMA described the failures as exposing consumers to “serious financial harm and lasting emotional distress,” noting the vulnerabilities went undetected for more than a year. In addition to the penalty, Southern Phone has entered a 36‑month court‑enforceable undertaking requiring an independent review of its systems, regular security testing, and ongoing reporting to ACMA as the regulator continues a broader crackdown on mobile‑number fraud.

ACMA’s wider crackdown has already led to over $4.6 million in fines across the sector this financial year. Southern Phone’s parent company, AGL, issued a public apology which included admitting to its failures, and pledged to enhance identity verification and upgrade internal systems to stop future breaches. Regulators are urging all telecommunications companies to strengthen their defences against increasingly advanced scams that target mobile number-based authentication.

The Office of the Australian Information Commissioner (OAIC) has commenced a targeted compliance sweep focusing on businesses that collect personal information face to face. This includes sectors such as property, hospitality, car rental, and retail, where customer details are often provided quickly and with limited explanation.

The regulator’s attention is firmly on transparency. Where personal information is collected at open homes, service counters, or venue entrances, individuals may not clearly understand how their information will be used, stored, or shared. The OAIC’s current activity is aimed at closing that gap between collection practices and customer understanding.

Organisations whose privacy policies do not accurately reflect their practices, or are difficult to access or understand, may face regulatory action. Recent reforms have reinforced the expectation that privacy policies are not generic documents but living statements that reflect what happens on the ground.

As part of this sweep, businesses should take the opportunity to review whether their privacy policy clearly:

• explains what personal information is collected through in-person interactions and why;
• describes how that information is used, stored, and ultimately disposed of;
• is easy to find and written in clear, customer-friendly language; and
• aligns with actual collection practices, particularly where staff collect information directly.

This initiative is a timely reminder that privacy compliance is not just a legal requirement. Clear and accurate privacy practices are increasingly central to customer trust and operational certainty.

Summary

In December 2025, the Australian Signals Directorate’s Australian Cyber Security Centre (ACSC) and the American Cybersecurity and Infrastructure Security Agency (CISA), in collaboration with various other international security and intelligence agencies, published guidance for the secure integration of artificial intelligence (AI) into operational technology (OT) for critical infrastructure assets (Guidance).

This Guidance (which can be found here) outlines four key principles that critical infrastructure owners and operator should follow to leverage the benefits of AI in OT while reducing risk.

Background

The functioning of Australia’s society and economy is underpinned by critical infrastructure assets. These assets, which include water, electricity and port assets, provide the wider community with access to essential services.

Most critical infrastructure in Australia is either privately owned and operated, or run on a commercial basis by government. The responsibility for ensuring the continuity of operations and the provision of essential services to the Australian economy and community is shared between owners and operators of critical infrastructure, and federal, state and territory governments.

Critical infrastructure assets make extensive use of OT. OT refers to programmable systems or devices that either interact with the physical environment, or manage devices that interact with the physical environment. These systems or devices detect or cause a direct change through the monitoring and/or control of devices, processes, and events. Examples of OT include industrial control systems, fire control systems, physical access control mechanisms, pressure monitors, and leak detectors.

The growth of AI presents an opportunity to critical infrastructure owners and operators to enhance the efficiency, productivity and decision-making of OT systems. However, the integration of AI into the OT systems of critical infrastructure assets also introduces new risks and challenges that require careful management to support the safety, security, and reliability of OT systems of critical infrastructure assets.

The Guidance

Against this backdrop, ACSC and CISA, in collaboration with various other international security and intelligence agencies, have published the Guidance.

The Guidance states that in order to successfully mitigate the risks of integrating AI into OT systems, it is essential for critical infrastructure owners and operators to follow the following four principles:

Understand AI

Understand the unique risks and potential impacts of AI integration into OT environments, the importance of educating personnel on these risks, and the secure AI development lifecycle.

Consider AI Use in the OT Domain

Assess the specific business case for AI use in OT environments and manage OT data security risks, the role of vendors, and the immediate and long-term challenges of AI integration.

Establish AI Governance and Assurance Frameworks

Implement robust governance mechanisms, integrate AI into existing security frameworks, continuously test and evaluate AI models, and consider regulatory compliance.

Embed Safety and Security Practices Into AI and AI-Enabled OT Systems

Implement oversight mechanisms to ensure the safe operation and cybersecurity of AI-enabled OT systems, maintain transparency, and integrate AI into incident response plans.

We outline each of these principles in more detail, below.

Principle 1

Principle 1 (Understand AI) includes:

• understanding the unique risks of AI and its potential impact to OT. The Guidance provides a broad overview of known AI risks in an OT environment that critical infrastructure owners and operators should consider, including cybersecurity risks, lack of explainability, AI dependency, reliability, and interoperability issues, impacts on OT, and possible mitigations. The risks listed in the Guidance are non-exhaustive, and it is recommended that critical infrastructure owners and operators investigate the risks specific to their organisation.
• understanding the secure AI system development lifecycle. Critical infrastructure owners and operators should verify that the AI system was designed securely and understand the roles and responsibilities through the AI system’s lifecycle. Owners and operators should clearly define and communicate these roles and responsibilities with the AI system manufacturer, OT supplier, and any system integrator or managed service provider roles. Critical infrastructure owners and operators should also carefully evaluate the trade-offs between different methods for sourcing an AI system (i.e., procuring vs. developing vs. customising an AI system). Where possible, critical infrastructure owners and operators should demand AI systems that are secure by design and will not negatively impact OT operation and safety.
• educating personnel on AI. Integrating AI into OT environments can lead to personnel relying too much on automation, resulting in dependency risks, skill erosion, and skill gaps. Critical infrastructure owners and operators may mitigate these risks by focusing on skill development and cross-disciplinary collaboration, such as training OT teams to interpret and validate AI outputs and maintain operational competencies alongside AI systems, developing clear standard operating procedures for all AI-related operations, interventions, and incidents, and having operators request that AI outputs include clear and transparent documentation of decision-making processes to help humans to better understand and validate outputs.

Principle 2

Principle 2 (Consider AI Use in the OT Domain) includes:

• considering the OT business case for AI use. Before incorporating an AI system into their OT environment, critical infrastructure owners and operators should consider whether an established capability meets their needs before pursuing more complex and novel AI-enabled solutions.
• managing OT data security risks for AI systems. When integrating AI into OT environments, critical infrastructure owners and operators should work with AI model developers to address data-related challenges, including data assurance (e.g., where OT data for training AI models is stored, who can view, access, or modify it, and how AI vendors access and use the OT data), and data privacy and security (e.g., by instituting protection from access abuse, intentional or inadvertent data poisoning, or dependency on synthetic, generated data).
• understanding the role of OT vendors in AI integration. OT vendors play a crucial role in advancing AI integrations into OT environments. Critical infrastructure owners and operators should demand transparency and controls from OT vendors regarding how AI technologies are embedded into their products.
• evaluating challenges in AI-OT system integration. When integrating AI into OT environments, critical infrastructure owners and operators should carefully evaluate the existing infrastructure to ensure compatibility and security. The guidance lists some challenges an organisation may face when integrating AI into OT systems, as well as recommending some strategies to help mitigate those challenges.

Principle 3

Principle 3 (Establish AI Governance and Assurance Frameworks) includes:

• establishing governance mechanisms for AI in OT. This involves establishing clear policies, procedures, and accountability structures for AI decision-making processes within OT.
• integrating AI into existing security and cybersecurity frameworks, including by embedding AI system assessments within existing risk evaluation, mitigation, and monitoring processes. This means that traditional cybersecurity requirements, vulnerability management, and critical infrastructure regulations must be factored in when integrating AI systems.
• conducting thorough AI testing and evaluation when introducing AI into OT environments to support the safe and reliable operation of these systems. Operators should only move the AI system into production for additional testing after sufficient testing in a non-production environment.
• navigating regulatory and compliance considerations for AI in OT. Critical infrastructure owners and operators should evaluate the applicability of current AI technical standards in their OT domain, continuously validate and verify that the performance of AI systems meets stringent OT performance and safety regulations, and identify and deploy thresholds for defaulting back to non-AI systems in OT, for example if AI system outputs fall below performance and safety thresholds.

Principle 4

Principle 4 (Embed Oversight and Failsafe Practices into AI and AI-Enabled OT systems) includes:

• establishing monitoring and oversight mechanisms for AI in OT. This includes establishing safety thresholds, alternative sensor output, or state changes that add human-in-the-loop intervention points, establishing safe operating bounds for OT devices that detect AI drift, model changes, or security risks, implementing key performance indicators that measure AI effectiveness, address concerns, and identify areas for improvement, and using push-based or brokered architectures that move required features or summaries out of OT without granting persistent inbound access.
• embedding safety and failsafe mechanisms. This includes establishing failsafe mechanisms that enable AI systems to fail gracefully without disrupting critical operations, designing functional safety procedures that account for the AI system, and incorporating AI considerations into the cybersecurity incident response plan.

Conclusion

The ACSC and CISA concludes that by adhering to the Guidance and continuously monitoring, validating, and refining AI models, critical infrastructure owners and operators can achieve a balanced integration of AI into the OT environments that control vital public services.

The National AI Plan (Plan) released on the 2^nd of December 2025 aims to set a strategic direction for artificial intelligence (AI) and provide a framework for ongoing action as technologies evolve. Structured around three core objectives the Plan aims to guide investment, adoption, and AI governance in a way that balances innovation with safety and inclusion.

Infrastructure and Capability

The first pillar focuses on building the foundations for AI at scale through investment in high-speed connectivity, sustainable data centres, and advanced computing power. This includes expanding the National Broadband Network, improving regional connectivity and ensuring nationwide access to AI-enabled services. These initiatives coincide with major global investments, such as Amazon committing $20 billion to data centre expansion, and Microsoft’s $5 billion in cloud and AI infrastructure. For industries that rely on technology and data, these changes mark a turning point as AI adoption accelerates across sectors, reshaping operational models and risk landscapes. Organisations will need to modernise workflows, strengthen cyber resilience, and consider ESG factors as growth becomes tied to sustainability and security expectations.

Spread the Benefits

The second strategy focuses on ensuring that every Australian can benefit from AI, regardless of demographic differences. Its aim is to build digital capability across society and reduce inequalities. The first priority is to scale AI adoption among SMEs, to drive innovation, employment, and productivity. To achieve this the government plans to support SMEs and not-for-profits through tailored guidance and resources provided by the National AI Centre and AI Adopt program, with particular attention to regional and disadvantaged communities.

The second priority is to strengthen skills and support workforce transitions by embedding AI capabilities across education and training systems, promoting lifelong learning and reducing digital literacy gaps. Collaboration with unions and employers will help ensure fair and inclusive adoption of AI in workplaces.

The final goal is to improve public services through responsible AI use, supported by appropriate human oversight, to enhance efficiency and accessibility. Planned initiatives include using AI in healthcare and schools, supporting environmental management efforts, and appointing a Chief AI Officer in every public sector agency to strengthen trust and transparency.

Keep Australians Safe

The third strategy focuses on protecting Australians by ensuring that AI is developed and used within strong and adaptable legal and regulatory frameworks. This includes reducing risks, promoting responsible practices, and shaping global norms for AI governance. To support this the government will update and enforce existing laws to address issues, such as bias and privacy and will establish an AI Safety Institute to monitor, test and advise on emerging threats. The strategy also introduces sector-specific safeguards in areas such as consumer protection, healthcare and copyright to ensure that AI technologies are used safely and ethically across the economy.

A further element of this strategy is to strengthen public confidence by encouraging responsible AI practices, including greater transparency around AI-generated content, and alignment with international standards and best practices. Australia will also work with international partners to shape global standards for AI governance, helping to ensure that emerging rules protect local interests while supporting ongoing innovation and competitiveness.

Key Takeaways

The Plan signals a shift in expectations for organisations across all sectors. As AI becomes embedded in decision-making, regulatory and consumer scrutiny will continue to rise and organisations will need to demonstrate fairness, transparency, and compliance. Insurers and brokers should be prepared to engage with regulators on the use of AI in pricing, claims-handling and underwriting, and to meet growing consumer expectations for clearer explanations and greater confidence in automated assessments.

Workforce adaptation will also be critical, supported by training and governance frameworks that help manage hybrid workflows and validate AI outputs. Reputational and legal risks are expected to increase as errors in AI-driven assessments or biased pricing could expose organisations to regulatory action and public backlash. Organisations that adopt best-practice AI governance early will be better positioned to manage emerging risks and maintain trust in an evolving digital landscape.

In late 2025, the Office of the Privacy Commissioner (OPC) issued its annual report on privacy regulation in New Zealand (here). The Report provides a summary of the OPC’s involvement over the past year, key achievements, its strategic direction and considerations under the Privacy Act 2020 (the Act). The OPC also took the opportunity to emphasise that New Zealand’s privacy legislation is falling behind, and reform of the Act is increasingly urgent.

An increase in complaints and breach notifications

The OPC reported a sharp uplift in both the volume and complexity of complaints and breaches notifications. In total, there were 1,598 complaints and 1,093 breaches notified over the reporting period. These numbers are a 27% increase from the year before and reflects a growing awareness of privacy obligations among agencies across New Zealand, alongside higher expectations for compliance and accountability.

Key achievements

The intangible nature of privacy makes it difficult to measure. Despite this, the Report sets out meaningful progress towards three core objectives outlined in the OPC’s Statement of Intent 2023-2027 (see here).

Clear expectations were set with the development and issuance of the Biometrics Code of Practice (see here). The Code came into force on 3 November 2025 and provides a legal framework for the use of biometric processing and is supplemented by detailed guidance to help agencies understand their responsibilities and enhance compliance.

The Māori Reference Panel was established this year to bring a te ao Māori perspective to privacy regulation. This aligns with its intention of partnership and ensuring that the OPC Codes of Practice and guidance are culturally appropriate and fit for purpose.

The Facial Recognition Technology trialled by Foodstuffs North Island (see here), was investigated by the OPC as a means to ensure appropriate privacy protections in an increasingly digital environment. While the OPC found that the trial was compliant with the Act, it used the investigation to outline considerations for agencies using FRT and reenforce their responsibilities and expectations for accountability.

Reform is increasingly urgent

To quote the report, “The Act is increasingly ‘long in the tooth’ as it is based on Law Commission recommendations from 2011”. A range of reforms are needed to ensure that New Zealand’s privacy and data protection framework remain fit for purpose. In particular the OPC is recommending four specific amendments:

• Inclusion of a “right to erasure”, providing individuals with the right to request that their data be deleted.
• A new and significantly stronger penalty regime.
• An accountability principle, requiring agencies demonstrate how they meet their privacy requirements.
• Stronger protections with regard to automated decision making.

Overall, a busy year for the regulator, and one which has both highlighted an increased engagement in privacy and also some of the shortcomings in our regulatory framework.

Singapore is strengthening board‑level cyber governance across critical infrastructure operators, with new requirements for board members in key sectors such as energy, healthcare and telecommunications to undergo mandatory cybersecurity training by early 2026, alongside enhanced oversight responsibilities for Critical Information Infrastructure (CII). These measures include granting chief information security officers’ direct access to boards and expanding accountability under the updated Cybersecurity Code of Practice, reflecting a rapidly evolving threat landscape and the government’s move to share classified threat intelligence with CII operators.

In Singapore, the Cyber Security Agency of Singapore (CSA), is the national authority responsible for securing Singapore’s cyberspace and protecting its CII. The implementation of the new CSA CII requirements establishes cybersecurity as a critical responsibility for boards, requiring directors to deepen their strategic risk understanding, strengthen governance frameworks, and for directors to assume a proactive role in resilience-focused decision-making. Complemented by initiatives such as the Cyber Resilience Guide for Boards, these developments represent a substantial advancement in Singapore’s approach to safeguarding vital digital infrastructure by cultivating informed, cyber-capable leadership.

With ransomware and other cyber incidents increasing across the region, Thailand’s Personal Data Protection Committee (PDPC) is adopting a more assertive enforcement posture. Its priorities include strengthening compliance with data breach notification requirements and ensuring organisations have timely reporting mechanisms and robust incident response protocols in place to safeguard personal information. This heightened scrutiny presents compliance challenges for organisations still developing their breach response capabilities. However, it also creates an opportunity for organisations that invest early in robust incident response frameworks and notification procedures to reduce regulatory risks and strengthen cyber resilience. Organisations that treat 2025-2026 as a window to build these capabilities, rather than waiting for an incident to occur, will be far better positioned to manage incidents and meet the PDPC’s growing oversight with confidence.

35‑year‑old Russian national Denis Obrezko was arrested in Phuket on 6 November 2025. Mr Obrezko, allegedly a member of the Kremlin‑linked hacking group “Void Blizzard,” was accused of breaching government systems across the U.S. and Europe using techniques such as password spraying, stolen authentication details and targeted spear‑phishing campaigns. Thai authorities, acting on an FBI tip, seized Mr Obrezko’s laptops, mobile phones and digital wallets in a coordinated operation against state‑aligned cyber‑espionage. Thai police initially withheld Obrezko’s name, but Russian state media quickly identified him and confirmed the arrest through family members.

Microsoft Threat Intelligence identifies Void Blizzard as an espionage‑focused group targeting NATO members, Ukraine, and critical sectors including transportation, healthcare and media (here and here). Obrezko is currently held in Bangkok awaiting extradition to the U.S., while Russian diplomats have confirmed consular contact but offered no public comment.