ESPNFollow live: Chile evens score vs. USMNTThe Jerusalem PostTrump, tech executives release 'morally binding' accord amid rising fears over AI safetyESPN DeportesPese a dominar, Estados Unidos se descuida y permite empate chilenoCNN TürkRESMİ GAZETE KARARLARI 30 EYLÜL: Resmi Gazete kararları neler oldu? İstanbul Beykent Üniversitesi Öğretim Üyesi alımı ilanı yayınlandı!PunchLagos exhibition creates market for artistsDaily MaverickRANKINGS UP: Cape Town hosts world’s higher education elite while SA universities riseInquirerMarcos institutionalizes DBM’s Compass for transparent gov’t spendingColliderMichael Fassbender’s R-Rated Sci-Fi Thriller Officially Loses Box Office MomentumUOLSuspeito em fuga pula muro de escola, assusta crianças e vai preso pela GCM em ApucaranaMalay MailFlash floods in major urban centres and cities are becoming too frequent in Asean — Phar Kim Beng매일경제“자연·교통·직장 3박자”…아크메르 동탄 10월 공급الشرقوسط تحقيقات مالية في تركيا.. أردوغان يشكل لجنة لإدارة تصفية صناديق استثمار
The Daily Newsstand · Free, Always
Wednesday, September 30, 2026

Government takes early action to protect ‘legacy’ technology systems against AI breaches

Translate

Acting Home Affairs Minister Richard Marles. Mick Tsikas/AAP

The Albanese government is strengthening its defences against rogue AI activity in the wake of the OpenAI breach of a government website and ahead of the report from a taskforce it established to investigate the incident.

A direction has been given to departments to “harden” cyber security defences on their “legacy” technology systems.

The Australian Signals Directorate reported last year that 59% of Australian government entities reported legacy technologies were impacting their ability to implement key cyber security controls.

The OpenAI agent was able to breach a legacy Services Australia portal containing Medicare statistics. The breach occurred in June but the government was only notified in September. The old site has since been closed.

Read more: Australia’s legacy systems were already a cyber risk. AI agents are raising the stakes

Under the new direction issued by the secretary for the Department of Home Affairs, departments and agencies must:

  • conduct an immediate stocktake of their legacy systems

  • set a target to reduce this technology, backed by a risk management plan, and

  • report what they’ve done to the Department of Home Affairs.

The action comes as OpenAI, in a long statement, apologised for what happened in Australia and promised to “do better”. It will send a senior member of its team to appear next week before the Australian parliamentary committee examining AI.

The taskforce the government set up to make recommendations following last week’s announcement of the breach will report within weeks.

The government said in a statement today that its “most critical systems” will be prioritised in the review process. The move “is designed to fortify against all forms of AI-enabled threats: inadvertent, or deliberate and targeted”.

The direction also requires agencies to “balance system availability with security, particularly where systems are public facing”.

This will ensure agencies consider requirements for rapid patching and security when managing critical government systems.

Acting Home Affairs Minister Richard Marles said:

AI is changing the environment in which we operate at extraordinary speed. Government systems need to keep up.

We can’t wait for an old system to fail before replacing it. We need to identify vulnerabilities and deal with them before they can be exploited.

View the original on The Conversation →

KioskNews shows a cleaned-up reading view extracted from the publisher’s page — the original always lives on their site, not ours.