ASD's cyber centre has issued a critical alert telling organisations to prepare for shorter patch timeframes, after some compromises came within 48 hours of a fix. A day later CISA urged immediate patching of Java4Shell, a Java flaw researchers say can expose Jira. Slow patching is a board risk.
Google’s new TPU 8tx and 8i Max target confidential inference and massive context workloads, intensifying the contest with Nvidia across AI infrastructure.
Jensen Huang’s departure puts Nvidia’s leadership and the sovereign AI ambitions of its customers under scrutiny. As governments build on its chips, the question is whether they understood the risk of relying so heavily on one company, and the founder whose judgement has shaped its future.
ASD warns of shrinking patch windows as CISA urges immediate Java4Shell fixes
ASD's cyber centre has issued a critical alert telling organisations to prepare for shorter patch timeframes, after some compromises came within 48 hours of a fix. A day later CISA urged immediate patching of Java4Shell, a Java flaw researchers say can expose Jira. Slow patching is a board risk.
Australia’s national cyber security agency is warning organisations to prepare for more frequent software updates and less time to apply them. On 17 November, the Australian Signals Directorate’s Australian Cyber Security Centre (ACSC) issued a critical-rated alert urging organisations to speed up patching, reduce the number of systems exposed to the internet and check that suppliers can promptly update the systems they manage.
The ACSC says it continues to respond to incidents in which attackers exploit unpatched software, sometimes combining several vulnerabilities to gain wider access, steal data or take control of systems. Its Annual Cyber Threat Report 2025–26 found that publicly reported software vulnerabilities rose 30 per cent over the financial year. In some cases, attackers exploited vulnerabilities within 48 hours of patches or mitigation advice becoming available. The ACSC says AI and automation are helping attackers find and exploit flaws faster and across more systems.
A separate US alert the same week illustrated the pressure. On 17 November US time, Oracle released fixes for CVE-2026-114938, a memory corruption flaw in Java’s Base64 decoder known as Java4Shell. Under certain conditions, it can let an attacker run code remotely without signing in. Oracle rated it 8.1 out of 10 under CVSS 3.1 and assessed it as difficult to exploit. Calif, the cyber security firm that discovered the flaw, rated it 9.2 under CVSS 4.0. Calif says its researcher used AI to find the flaw in about an hour and develop an initial working exploit in another two hours, spending less than US$200 on AI usage.
Later that day, the US Cybersecurity and Infrastructure Security Agency (CISA) urged organisations to apply available fixes immediately, starting with internet-facing systems. It also advised them to look for Java runtimes bundled inside third-party products, container images and appliances, where they may not appear as separate entries in software inventories.
Calif demonstrated unauthenticated remote code execution in lab tests against default installations of Keycloak, self-managed Atlassian Jira and Confluence, JFrog Artifactory, Jenkins and PingFederate. The researchers said exploitation depends on the individual application. The flaw affects certain unpatched OpenJDK 17 builds from 17.0.6 onward and affected builds in the Java 21, 25 and 27 lines. It requires compatible x86-64 hardware and hardware-accelerated Base64 decoding. ARM64 systems are unaffected. Oracle also lists a specific Java 8 performance build and several GraalVM builds.
The pressure extends beyond this Java flaw. On 13 November, the Cyber Security Agency of Singapore said the average monthly number of reported cyber incidents since September was three times the January to August average. In its November System Risk Outlook, the Australian Prudential Regulation Authority (APRA) warned that advances in AI could give financial institutions and their service providers less time to assess their exposure, apply patches and contain attacks.
“Australia’s financial system remains resilient, but the risk environment is becoming less forgiving.” - John Lonsdale, Chair of the Australian Prudential Regulation Authority
Why does it matter?
After an incident, regulators, insurers and customers may ask how quickly an organisation acted once a fix or mitigation became available. The ACSC specifically advises organisations to confirm that suppliers can promptly patch the systems they manage. For boards, that makes patch speed a question about supplier contracts and service commitments as well as internal processes.
The ACSC recommends enabling automatic updates where appropriate, reducing unnecessary internet exposure, strengthening access controls with multi-factor authentication and addressing legacy or unsupported systems. Where an update has to wait, it advises assessing the risk, restricting exposure and putting temporary controls in place until the fix can be applied.
For Java4Shell, security teams should inventory their Java runtimes, including those bundled with vendor products or deployed in containers. They should prioritise internet-facing systems, confirm which build contains the fix and restart affected applications after updating so they load the patched runtime. Where an immediate update is unavailable, CISA recommends limiting unauthenticated access as one temporary measure.
Boards can ask their technology teams and managed service providers two questions this week. How long does it take to identify and patch every exposed system after a critical update is released? What controls can be put in place while a fix is pending?
Get the stories that matter to you. Subscribe to Cyber News Centre and update your preferences to follow our Daily 4min Cyber Update, Innovative AI Startups, The AI Diplomat series, or the main Cyber News Centre newsletter — featuring in-depth analysis on major cyber incidents, tech breakthroughs, global policy, and AI developments.
Sign up for Cyber News Centre
Where cybersecurity meets innovation, the CNC team delivers AI and tech breakthroughs for our digital future. We analyze incidents, data, and insights to keep you informed, secure, and ahead.
AI is no longer merely writing phishing emails. It is probing networks, exploiting access and directing intrusions at scale. As investigations expose both malicious agents and runaway security tools, businesses face a defining question: who controls the software acting on their behalf in real time?
AI is finding flaws faster than defenders can patch them, while autonomous agents push past human instructions. From exposed government systems to an Anthropic model that ignored stop signs, the question is who can halt the machine—and who pays when it acts without permission or moves the money.
Hot inflation, record AI credit stress and a seven-day chip slide showed that the boom’s limit is no longer demand but the price of money. Investors are rotating rather than fleeing, and the bill is landing far from Silicon Valley.
OpenAI flew its chief strategy officer to Sydney to apologise for a rogue AI agent's Medicare breach. The hack was clumsy. The silence that followed was worse, and Parliament now has to decide whether sorry is the only accountability on offer.
Where cybersecurity meets innovation, the CNC team delivers AI and tech breakthroughs for our digital future. We analyze incidents, data, and insights to keep you informed, secure, and ahead. Sign up for free!