Review mobile device settings
Check that mobile device management keeps work and personal apps and data apart, runs an always-on VPN, blocks USB data transfers and requires password-based lock screens.

ASD's eight prioritised mitigation strategies, at three maturity levels.
152 requirements

The Australian Government's Information Security Manual, guideline by guideline.
1,143 controls

Everything ASD added, changed and rescinded in the March 2026 Information Security Manual, in plain English. Open any control for its full guidance.
Reviewed 15 November 2026 · Compared with the December 2025 ISM · 1,081 controls in force
ASD added 9 controls, changed 12 and rescinded 1. Mobile devices and connected vehicles get new rules, and system owners now plan for resilience as well as security.
The release brings the ISM to 1,081 controls. Mobile devices keep work and personal apps and data apart, run an always-on VPN and block data transfers over USB. Three controls limit what staff do in and near connected vehicles. System owners now set resilience objectives and manage, not just monitor, cyber threats, security risks and controls. Four emanation security controls move to ASD's new terms, and five other controls are reworded only. ISM-1837, on password never expires and password not required settings, was rescinded.
What this release means if you assess, run or build systems against the ISM.
Check that mobile device management keeps work and personal apps and data apart, runs an always-on VPN, blocks USB data transfers and requires password-based lock screens.
System owners now set resilience objectives alongside security objectives, based on the impact if a system were compromised or attacked, and tailor controls to both.
Keep mobile devices off connected vehicles' infotainment systems. In or near a connected vehicle, do not view sensitive or classified data on a mobile device or hold sensitive or classified calls and conversations.
Organisational data that AI applications generate, collect or process is used to train, fine-tune or improve AI models only with its data owners' informed, explicit consent, obtained in advance.
Grouped by theme. Open a control for its plain-English guidance, steps and audit checks.
Mobile devices keep organisational and personal apps and data apart, run an always-on VPN and block data transfers over USB. Staff using privately owned mobile devices or desktop computers to access OFFICIAL: Sensitive or PROTECTED systems or data cannot give unapproved AI agents access.
Mobile devices never connect to a connected vehicle's infotainment system. In or near connected vehicles, sensitive or classified data is not viewed on mobile devices, and sensitive or classified calls and conversations are not held.
Existing software artefacts in the authoritative source for software are tested periodically for known weaknesses, using SAST, DAST or SCA to suit each artefact type. Organisational data that AI applications generate, collect or process is used to train, fine-tune or improve AI models only with its data owners' informed, explicit consent, obtained in advance.
Controls whose requirement changed. Open "Compare ASD's wording" to see the statement with removed words struck through and added words highlighted.
Four controls use ASD's new emanation security terms.
Now an emanation security risk assessment, not a threat assessment.
When an emanation security Removed: threat Added: risk assessment is required, it is sought as early as possible in a system's life cycle.
Now an emanation security risk assessment, not a threat assessment.
System owners deploying SECRET or TOP SECRET systems in mobile platforms, or as a deployable capability, contact ASD for an emanation security Removed: threat Added: risk assessment.
Now an emanation security risk assessment, not a threat assessment.
System owners deploying SECRET or TOP SECRET systems within fixed facilities contact ASD for an emanation security Removed: threat Added: risk assessment.
Now refers to emanation security mitigation advice instead of TEMPEST requirements statements.
Recommended actions contained within Removed: TEMPEST requirements statements Added: emanation security mitigation advice issued for systems are implemented by system owners.
System owners now manage risks as well as monitoring them, and plan for resilience.
System owners continuously monitor each system's security and now also manage its cyber threats, security risks and controls.
System owners Added: continuously monitor Added: the security of each system, and Added: manage associated cyber threats, security risks and Removed: controls, on an ongoing basis.Added: controls.
Adds resilience objectives, based on the impact if the system were compromised or attacked.
System owners, in consultation with each system's authorising officer, determine the system boundary, business Removed: criticality Added: criticality, and security Added: and resilience objectives for each system based on an assessment of the impact if it were to be Removed: compromised.Added: compromised or attacked.
Controls are tailored to resilience objectives as well as security objectives.
System owners, in consultation with each system's authorising officer, select controls for each system and tailor them to achieve desired security Added: and resilience objectives.
Internet access through the gateway no longer names a VPN, and mobile lock screens must be password-based.
Mobile devices and desktop computers still reach the internet through the organisation's gateway, but the control no longer names a VPN connection, so other technical solutions can meet it. Mobile devices still need always-on VPN functionality under the new ISM-2097.
Mobile devices and desktop computers access the internet via Removed: a VPN connection to an organisation's internet gateway rather than via a direct connection to the internet.
Lock screens must be password-based.
Mobile devices are configured with secure Added: password-based lock screens.
Testing now covers every software artefact, using the method that suits it, and scanning looks for malicious content, not only malicious code.
Scanning now looks for malicious content, not only malicious code.
All software artefacts are scanned for malicious Removed: code Added: content before being imported into the authoritative source for Removed: software, including all compiled code, third-party libraries and software components.Added: software.
Covers all software artefacts, not only third-party ones, and uses SAST, DAST or SCA to suit each artefact type rather than all three. Periodic retesting moved to the new ISM-2102.
All Removed: imported or referenced third-party software artefacts are tested Added: to detect known weaknesses using static application security testing (SAST), dynamic application security testing (DAST) Removed: and Added: or software composition analysis Removed: (SCA) Added: (SCA), depending on the software artefact type, before being imported into the authoritative source for Removed: software and periodically throughout the software development life cycle.Added: software.
When ML-DSA and ML-KEM are used, the preferred prerequisite is now FIPS 140-3 validation.
Now prefers FIPS 140-3 validation, instead of adherence to prerequisite FIPS publications.
When using ML-DSA and ML-KEM, as per FIPS 204 and FIPS 203 respectively, adherence to pre-requisite FIPS Removed: publications Added: 140-3 validation is preferred.
Changes ASD lists as wording only, keeping each control's intent. Check any policy or plan that quotes these controls, and compare the wording wherever it is shown.
System owners obtain Added: an authorisation to operate Added: for each non-classified, OFFICIAL: Sensitive, PROTECTED and SECRET system from its authorising Removed: officer based on the acceptance of the security risks associated with its operation.Added: officer.
System owners obtain Added: an authorisation to operate Added: for each TOP SECRET system, including Added: for each sensitive compartmented information system, from Director-General ASD (or their Removed: delegate) based on the acceptance of the security risks associated with its operation.Added: delegate).
Personnel Removed: accessing OFFICIAL: Sensitive or PROTECTED systems or data using privately-owned mobile devices or desktop computers Added: to access OFFICIAL: Sensitive or PROTECTED systems or data have enforced separation of classified data Removed: from Added: and personal data.
Personnel Removed: accessing systems or data using Removed: an organisation-owned mobile Removed: device Added: devices or desktop Removed: computer Added: computers to access classified systems or data have enforced separation of classified data Removed: from Added: and personal data.
Personnel Removed: accessing OFFICIAL: Sensitive or PROTECTED systems or data using privately-owned mobile devices or desktop computers Added: to access OFFICIAL: Sensitive or PROTECTED systems or data are prevented from storing classified data on their privately-owned mobile devices and desktop computers.
No longer part of the ISM. Each keeps its page, marked as rescinded, for anyone still working to an earlier release.
Rescinded. ASD names no replacement in its change notes.
The statement as ASD last published it
User accounts are not configured with password never expires or password not required.
44 new23 changed2 rescinded82 reworded
44 new controls covering AI agents, application and workload credentials, OAuth consent, session tokens and MACsec, and three yearly tasks move to every six months.Read the September 2026 changesJune 202620 new18 changed101 reworded
AI-assisted security testing and monitoring, advice for staff on posting about work online, and wider isolation and encryption controls. Nothing rescinded.Read the June 2026 changesDecember 202521 new9 changed6 rescinded68 reworded
Password rules follow current advice, a dozen AI controls arrive and fax messaging ends. Six controls rescinded, five of them about fax.Read the December 2025 changesSource: ASD's Information Security Manual, March 2026, compared control by control with the December 2025 release. Control statements are quoted as published. Everything else on this page is written by Control Stack.