

Cyber Security Action Month 2026: Website Checklist
Cyber Security Action Month began on 1 October 2026 with a simple message for Australians: take a second and stay secure. For business websites, the useful response is not another awareness poster. It is a short, evidence-based review of the systems that publish content, accept customer data, process payments, connect to other platforms and keep the site online.
The Australian Signals Directorate's Australian Cyber Security Centre (ACSC) is asking organisations to act on legacy technology, event logging and secure-by-design products and services. Its wider campaign advice also reinforces three controls that matter directly to website operations: updates, backups and multi-factor authentication.
This checklist translates those priorities into practical work for an Australian small or medium business. It covers the public website, CMS, hosting, domain and DNS, third-party integrations, deployment pipeline, forms, data and support arrangements. The goal is not perfect security in a month. It is to find the gaps most likely to turn a routine website problem into an outage, data breach or expensive recovery.
Why this topic is timely in October 2026
The 2026 campaign is aimed at individuals, SMEs, larger organisations and government. The ACSC says cyber threats are increasing in scale, speed and sophistication and that awareness must become year-round action.
That timing matters for websites because responsibility is usually fragmented. Marketing owns the CMS. A managed host controls infrastructure. An agency deploys code. Finance controls the payment account. A former employee may still own a plugin licence or DNS login. Each individual system can look acceptable while the connections between them remain unowned.
Cyber Security Action Month is a useful trigger for one joined-up review. Ask one business question: could we prevent, detect, contain and recover from a website incident with the people and access we have today? If the answer depends on an absent contractor, an unknown login or an untested backup, the business has found a concrete priority.
A practical website security checklist
Review the complete operating system around the website, not only the pages visitors can see.
Ownership
Name the business owner and technical owner for the website, domain, hosting, CMS, integrations and incident decisions.
Supported software
Inventory versions, dependencies and end-of-support dates across the CMS, application, server and delivery stack.
Privileged access
Remove stale accounts, minimise administrator rights and enforce MFA on every control plane that can change the site.
Recoverable backups
Protect code, content, data and configuration, then prove they can restore a working service within the business target.
Logging and alerts
Keep useful records for sign-ins, changes, deployments, application errors, edge activity and integration failures.
Integration security
Inventory forms, APIs, webhooks, payment services and secrets; restrict data and permissions to what each connection needs.
Secure change
Use review, testing, staged deployment and rollback for website releases instead of editing production directly.
Incident readiness
Document contacts, containment choices, evidence, communications, privacy assessment, recovery and post-incident review.
1. Map the website estate and assign ownership
Start with the systems that can affect availability, content, customer data or reputation. The list usually includes the domain registrar, DNS, CDN or web application firewall, hosting, operating system, runtime, database, CMS, code repository, deployment service, analytics, forms, email delivery, CRM, payments, search and backup service.
For each system, record the business owner, technical owner, support provider, renewal date, privileged accounts, data handled, dependencies and recovery contact. Add a clear escalation route for incidents outside business hours. If a supplier manages a service, confirm what the contract actually covers: patching, monitoring, backups, restoration, incident investigation and customer communication are different responsibilities.
This register is the foundation for every later control. Without it, teams can patch the CMS but miss the unsupported runtime beneath it, or secure hosting while leaving the domain registrar account exposed.
2. Patch the whole stack and plan out legacy technology
The ACSC describes unsupported technology as a significant and enduring risk because it no longer receives security updates. It can also give an attacker a path into newer connected systems.
For a website, patching must cover more than the CMS. Check:
- CMS core, plugins, modules, packages and custom extensions;
- application runtime and frameworks such as .NET, PHP, Node.js or Java;
- operating system, web server, database and container base images;
- deployment actions, build tools and package-lock changes;
- CDN, WAF, DNS, email, form and integration components;
- end-of-support dates for every major platform version.
Use automatic updates where the change is low risk and reversible. For production website code, pair prompt patching with a controlled process: take a recovery point, review release notes, build in a clean environment, run regression tests, deploy in stages and verify the live service.
If a platform cannot be upgraded immediately, document the exception, owner and deadline. Temporary measures can include tighter access, segmentation, application-surface reduction, additional logging and stronger account controls. A mitigation is a bridge to replacement, not a permanent version strategy.
3. Put MFA around every route that can change the website
The ACSC calls multi-factor authentication one of the most effective ways to protect valuable accounts and information. For websites, apply it to the CMS and to the less visible control planes that can bypass the CMS entirely.
Your coverage should include hosting, domain registrar, DNS and CDN, source control, deployment, cloud administration, database tools, backup consoles, payment administration, analytics, email delivery and vendor support portals. Prefer phishing-resistant options such as passkeys or hardware security keys where the service supports them, particularly for domain, cloud and source-control administrators.
Then review account hygiene. Remove former staff and agencies, replace shared administrator accounts with named identities, reduce standing privileges, store emergency access securely and test the recovery process for lost factors. A strong MFA policy is not complete if the only recovery email belongs to a former employee.
4. Back up what the website needs—and prove the restore
The ACSC recommends regular, protected backups and explicitly advises testing that they can be restored. That last step separates a backup record from a recovery capability.
A recoverable website may require:
- application code and exact dependency versions;
- CMS content, media and database records;
- environment and infrastructure configuration;
- DNS, CDN, WAF and email settings;
- integration mappings and webhook configuration;
- a secure method to replace secrets and keys;
- deployment instructions and provider contacts.
Define two business targets: how much recent data the organisation can afford to lose, and how long the website can be unavailable. Align backup frequency and restore design with those targets. Keep recovery copies isolated from the production account where appropriate so one compromised administrator cannot delete both the site and its backup.
Run a restore exercise into a separate environment. Verify pages, forms, authentication, integrations, media, redirects and scheduled jobs. Record the actual recovery time and every manual dependency the test exposes.
5. Keep logs that answer business questions
The ACSC says centralised event logging improves visibility and supports faster threat detection and incident response. The useful question for an SME is not how many logs it can store. It is whether the available evidence can explain what changed, who did it, what data or services were affected and whether the problem is still active.
Prioritise records for administrator sign-ins, failed access, permission changes, CMS publishing, plugin or package changes, deployments, WAF decisions, server and application errors, form abuse, API authentication failures, payment or CRM delivery failures and backup results.
Send important alerts somewhere a responsible person will see them. Define who responds, what counts as urgent and what the first checks are. Retention should be long enough to investigate incidents that are not discovered immediately, and access to the logs should be restricted so an attacker cannot quietly erase the evidence.
Monitoring also needs a business layer. A website returning HTTP 200 can still be failing if enquiries no longer reach the CRM, checkout callbacks are rejected or confirmation emails stop sending.
6. Review forms, APIs and connected business systems
Every integration expands the website's trust boundary. Inventory public forms, webhooks, CRM connections, booking systems, payment services, marketing platforms, analytics, file uploads and internal APIs.
For each connection, confirm:
- the business purpose and current owner;
- the minimum personal data and permissions required;
- where credentials are stored and how they are rotated;
- how requests are authenticated, validated, rate-limited and logged;
- what happens when the receiving system is slow or unavailable;
- how failed or duplicated transactions are reconciled;
- whether old API keys, test endpoints and former vendors have been removed.
Never place server credentials in browser code or public repositories. Treat uploaded files as untrusted. Validate input on the server, use least-privilege service identities and make high-value actions idempotent where retries are possible. VaniTech's inactive API token cleanup guide provides a focused process for credentials that no longer have a clear owner.
7. Buy and build for secure operation
Secure by Design means considering cyber threats throughout design, development, deployment and operation. Secure by Default means a product starts from a useful security baseline, with protections such as MFA, auditing and event logging available without extensive extra configuration.
Use those principles when selecting a CMS, host, ecommerce platform, plugin or support provider. Ask whether security updates are included, how quickly critical patches are applied, whether MFA and audit trails cost extra, how backups are isolated, how incidents are reported, what happens at end of support and how the business can export its data and configuration.
Apply the same standard to custom development. Require peer review, dependency scanning, secret management, separate environments, regression tests, staged deployment and rollback. Keep production changes attributable to a person or automated release. Avoid editing live code through a hosting file manager unless it is a controlled emergency procedure.
8. Prepare for a website incident before one happens
The ACSC says organisations should maintain and test a cyber security incident response plan. Its executive guidance asks whether critical systems and data are known, providers have response obligations, the organisation can detect an incident, recovery resources are accessible and legal and communication responsibilities are understood.
For a website incident, the plan should name who can:
- disable accounts, rotate credentials and block malicious traffic;
- take the site or a feature offline when continued operation creates more risk;
- preserve logs, configuration and other evidence;
- restore from a known-good state and verify connected systems;
- assess personal information exposure and reporting obligations;
- approve messages to customers, staff, regulators and suppliers.
The OAIC recommends a written, tested data breach response plan with clear roles and a strategy to contain, assess and manage an incident. Link that privacy process to the technical plan so the website team does not treat data exposure as only a server problem.
Run a short tabletop exercise: a CMS administrator account is compromised, enquiry data may have been accessed and the website begins redirecting visitors. Walk through the first hour, first day and recovery. The gaps in access, evidence, decisions and communication will become obvious quickly.
A 30-day action plan for an Australian SME
Week 1: Establish control
Name the website owner and technical owner. Build the system and supplier register. Confirm access to the domain, DNS, hosting, CMS, source code, deployment, backups and critical integrations.
Week 2: Close immediate gaps
Remove stale accounts, enable MFA, rotate exposed or ownerless secrets, apply high-priority security updates and identify unsupported components. Document any upgrade exception with an owner and deadline.
Week 3: Prove detection and recovery
Centralise the most useful logs, route alerts to a responsible person and run a backup restore into a separate environment. Test a real enquiry, transaction or booking from the website through to the receiving business system.
Week 4: Exercise the response
Run a tabletop incident, update provider contacts and communication roles, fix the most serious exercise findings and create a recurring quarterly review. Security work becomes sustainable when it has named owners, evidence and dates—not when it depends on someone remembering.
Sources Checked
- ACSC: Take a second to stay secure this Cyber Security Action Month
- ACSC: How to update your device and software
- ACSC: How to back up your files and devices
- ACSC: Multi-factor authentication
- ACSC: Legacy technology management
- ACSC: Event logging
- ACSC: Secure by Design
- ACSC: Cyber security incident response planning—executive guidance
- OAIC: Preparing a data breach response plan