Regulatory Map

Phase 2 SMS Codes — Australian Online Safety Act

Schedule 4 (Core Features) · Schedule 4A (Messaging Features) · Head Terms
◆

Categories of Material Captured

Head Terms §2.1
Class 1C Material CLASS 1 SUB
Class 1 material depicting specific fetish practices or fantasies. Excludes class 1A material (CSEM, pro-terror, extreme crime & violence).
✓ Text ✓ Images ✓ Video ✓ Audio
Head Terms §2.1 · Inherits broad RC classification framework
Class 2A Material CLASS 2 SUB
X 18+ film material depicting actual (not simulated) sexual activity between consenting adults. Includes deepfakes depicting actual sexual activity.
✗ Text ✓ Images ✓ Video ✗ Audio
Head Terms §2.1 · Film pathway only; X 18+ has no publication equivalent
Class 2B Material CLASS 2 SUB
Class 2 material depicting high-impact sexually explicit images (including high-impact nudity). Excludes class 2A. May include deepfakes.
✗ Text ✓ Images ✓ Video ✗ Audio
Head Terms §2.1 · "depicts… images" anchors to visual material
Self-Harm Material CLASS 2 SUB
Class 2 material that encourages, promotes or provides instruction for suicide, deliberate self-injury, or eating disorders / associated behaviour.
✓ Text ✓ Images ✓ Video ✓ Audio
Head Terms §2.1 · Function-based: "encourages, promotes or provides instruction"
High-Impact Violence Material CLASS 2 SUB
Class 2 material depicting shocking, gratuitous or exploitative real (or presented-as-real) images of violence against people/animals and/or gore.
✗ Text ✓ Images ✓ Video ✗ Audio
Head Terms §2.1 · "depicts… real images" anchors to visual
Violence Instruction Material CLASS 2 SUB
Class 2 material that encourages, promotes or provides instruction for physical violence intended to cause grievous bodily harm or kill another person.
✓ Text ✓ Images ✓ Video ✓ Audio
Head Terms §2.1 · Function-based: "encourages, promotes or provides instruction"
High-Impact Sexually Explicit Material CLASS 2 SUB
Class 2 material which describes sexual activity in a high impact manner.
✓ Text ✓ Images ✓ Video ✓ Audio
Head Terms §2.1 · "describes" captures text
Simulated Gambling Material CLASS 2 SUB
Computer game containing simulated gambling, classified or would be classified R 18+ under the Classification Act.
✗ Text ✗ Images ✗ Audio ✓ Computer Game
Head Terms §2.1 · Form-specific: computer games only
Composite Groupings Used in the Codes
Online Pornography
Class 1CClass 2AClass 2B
Head Terms §2.1 — "online pornography"
High-Impact Online Pornography
Class 1CClass 2A
Head Terms §2.1 — "high-impact online pornography"
High-Priority Restricted Category
High-Impact Online PornographySelf-Harm Material
Head Terms §2.1 — "high-priority restricted category of material"
Generative AI Restricted Category
Online PornographyHigh-Impact Sexually ExplicitSelf-HarmHigh-Impact ViolenceViolence Instruction
Head Terms §2.1 — "generative AI restricted category"
▲

Risk Profile Tiering

Schedule 4 Only · cl 4 & 5
Risk tiering applies only under Schedule 4 (Core Features) — it does not apply to Schedule 4A (Messaging Features), where all compliance measures apply uniformly. Under Schedule 4, the risk tier determines which compliance measures a service must meet. A provider may hold different tiers for different categories of material.
TIER 1High Risk
Highest compliance burden. Triggered where the risk that Australian children will access or be exposed to the relevant material is assessed as high, or where the provider auto-assigns Tier 1 under Head Terms §5.2(a)(ii). Requires age assurance (AI chatbot), on-platform reporting, complaints tools, pre-change assessments, plus all Tier 2 measures.
TIER 2Moderate Risk
Moderate compliance burden. Requires detection & removal systems (material not allowed), safety-by-design defaults (AI chatbot), terms & conditions, reporting mechanisms, training, eSafety reporting, and transparency measures.
TIER 3Low Risk
Lowest compliance burden. No specific tier-dependent measures under §§8 or 10. Providers still comply with §7 (material allowed) and §9 (supporting measures). Automatically assigned to enterprise/limited-feature services under cl 4.4.
Compliance Pathways & Risk Assessment
Pathway A — Material Allowed Under ToS
Where material is allowed, no risk assessment or tiering is required. Provider proceeds to §7 measures (age assurance, safety tools, publishing info, annual reporting) plus §9 supporting measures.
Sch 4, cl 4.2(a)(i) → Section 7 + Section 9
Pathway B — Material Not Allowed Under ToS
Where material is not allowed, the provider must undertake a risk assessment (or auto-assign Tier 1). The resulting tier determines §8 measures (detection & removal, eSafety reporting). §9 supporting measures also apply.
Sch 4, cl 4.2(a)(ii) → Risk assessment → Section 8 + Section 9
Pathway C — AI Companion Chatbot Features
AI chatbot features are risk-assessed separately per restricted category. If the chatbot's sole/predominant purpose is to generate a category, it is automatically Tier 1 for that category. §10 measures apply based on tier.
Sch 4, cl 4.1 → Risk assessment → Section 10 + Section 9
Deemed Tier 3 Exemptions
Certain services are deemed Tier 3 without risk assessment: (a) enterprise services limited to employees/customers; and (b) services without profile/connection features.
Sch 4, cl 4.4
Schedule 4A — No Tiering: All compliance measures in Schedule 4A apply uniformly to every messaging feature. No risk assessment, no tiering framework, and no tier-dependent differentiation.
▣

Compliance Measure Obligations

Schedules 4 & 4A
Schedule 4 — Core Features
Schedule 4A — Messaging (no tiering)
Click any obligation for details
1
Age Assurance & Access Controls
Age assurance & access controls — material allowed Material Allowed
Sch 4, cl 7.1
Online pornography & self-harm material where allowed under ToS.
Age assurance & access controls — AI chatbot Tier 1
Sch 4, cl 10.1
Before access to AI feature or generating restricted material.
Safety-by-design defaults — AI chatbot Tier 2
Sch 4, cl 10.2
Age assurance or prevent generation of restricted material.
2
Content Detection & Moderation
Detect & remove online pornography Tier 1/2
Sch 4, cl 8.1
Where material not allowed under ToS.
Detect & remove self-harm material Tier 1/2
Sch 4, cl 8.2
Where material not allowed under ToS.
Detect & remove high-impact violence material Tier 1/2
Sch 4, cl 8.3
Where material not allowed under ToS.
3
Safety Tools & Settings
Safety tools for end-users Material Allowed
Sch 4, cl 7.2
Filtering, blocking, blurring, halting autoplay. Default for children.
Tools, features & settings — messaging No Tier (4A)
Sch 4A, cl 10.5
Block/prevent unsolicited material. Default most restrictive for children.
Improvement of tools over time — messaging No Tier (4A)
Sch 4A, cl 10.8
Develop/improve tools. Engage in improvement activities annually.
4
Terms & Conditions
T&Cs for class 1C & class 2 material — core All Tiers
Sch 4, cl 9.1
Plain language. Graduated, risk-based enforcement.
T&Cs for generative AI restricted material Tier 1/2
Sch 4, cl 10.3
Clear policies on AI generation of restricted material.
T&Cs prohibiting sharing online porn to children No Tier (4A)
Sch 4A, cl 10.1
Prohibit sharing online porn by end-user to Australian child.
5
User Reporting Mechanisms
Reporting tools — core features All Tiers
Sch 4, cl 9.3
Accessible. Reporter identity protected. Evaluated & actioned.
On-platform reporting tools — core features All Tiers
Sch 4, cl 9.4
Available at the point users access posted materials.
Reporting tools — AI chatbot Tier 1/2
Sch 4, cl 10.4
Report AI-generated restricted material contrary to ToS.
On-platform reporting tools — AI chatbot Tier 1
Sch 4, cl 10.5
Visible at the point end-user generates materials.
Reporting mechanism — messaging No Tier (4A)
Sch 4A, cl 10.2
Report breaches of online-porn-to-child prohibition.
6
Complaints Handling
Complaints tools — core features All Tiers
Sch 4, cl 9.5
Complaints about report handling & Code compliance.
Inform users of report/complaint outcomes All Tiers
Sch 4, cl 9.6
Acknowledge & inform in reasonably timely manner.
Complaints tools — AI chatbot Tier 1
Sch 4, cl 10.8
Complaints about AI-generated restricted material handling.
Complaints tools — messaging No Tier (4A)
Sch 4A, cl 10.16
Complaints about Code breach re messaging feature.
Referral of unresolved complaints to eSafety — core All Tiers
Sch 4, cl 9.9
Material non-compliance complaints unresolvable in reasonable time.
Referral of unresolved complaints to eSafety — messaging No Tier (4A)
Sch 4A, cl 10.17
Promptly refer material non-compliance complaints to eSafety.
7
Personnel, Training & Review
Trust & safety function — core All Tiers
Sch 4, cl 9.2
Sufficient personnel with skills, experience & qualifications.
Trust & safety function — messaging No Tier (4A)
Sch 4A, cl 10.14
Sufficient personnel to oversee messaging feature safety.
Training — reports & complaints (core) All Tiers
Sch 4, cl 9.7
Trained in policies & procedures for reports & complaints.
Training — AI chatbot personnel Tier 1/2
Sch 4, cl 10.9
Trained in AI-related reports & complaints policies.
Training — messaging reports personnel No Tier (4A)
Sch 4A, cl 10.3
Trained in messaging report policies & procedures.
Review of reporting systems — core All Tiers
Sch 4, cl 9.8
Annual review of effectiveness.
Review of reporting mechanism — messaging No Tier (4A)
Sch 4A, cl 10.4
Annual review of effectiveness.
8
Transparency & User Information
Publish info about safety tools — core Material Allowed
Sch 4, cl 7.3
Clear info on tools to limit exposure.
Info about risk of harm to children — core All Tiers
Sch 4, cl 9.14
Publish actions to reduce harm risk to children.
Info about AI restricted material actions Tier 1/2
Sch 4, cl 10.6
Publish actions to reduce AI-generated harm to children.
Info about tools & contact mechanisms — messaging No Tier (4A)
Sch 4A, cl 10.9
Clear info about tools & reporting mechanisms.
Info to assist end-users managing risks — messaging No Tier (4A)
Sch 4A, cl 10.11
Steps end-users can take to manage & mitigate risks.
eSafety role & complaints info — core All Tiers
Sch 4, cl 9.12
eSafety role/functions & how to complain.
eSafety role & complaints info — messaging No Tier (4A)
Sch 4A, cl 10.10
eSafety role/functions & how to complain.
Dedicated online safety info location — core All Tiers
Sch 4, cl 9.13
Code info, third-party counselling contacts.
Dedicated online safety info location — messaging No Tier (4A)
Sch 4A, cl 10.12
Code info, counselling contacts.
9
eSafety Reporting & Engagement
Annual Code report — material allowed Material Allowed
Sch 4, cl 7.4
Steps taken & explanation of appropriateness.
Code report on request — material not allowed Tier 1/2
Sch 4, cl 8.4
On eSafety written request. Within 2 months.
Code report on request — messaging No Tier (4A)
Sch 4A, cl 10.13
On eSafety written request. Within 2 months.
Updates to eSafety on tech changes — core All Tiers
Sch 4, cl 9.10
Significant changes affecting child access/exposure.
Updates to eSafety on tech changes — AI chatbot Tier 1/2
Sch 4, cl 10.7
Significant changes affecting child generation risk.
Updates to eSafety on tech changes — messaging No Tier (4A)
Sch 4A, cl 10.6
Significant changes affecting child access/exposure.
Engagement with safety organisations — core All Tiers
Sch 4, cl 9.11
Engage with civil society, academics & government.
Engagement or collaboration — messaging No Tier (4A)
Sch 4A, cl 10.15
Engage with organisations or collaborate with industry.
Timely response to eSafety — messaging No Tier (4A)
Sch 4A, cl 10.18
Policies ensuring timely response to eSafety.
10
Pre-Change Assessments
Assessment before material changes — AI chatbot Tier 1
Sch 4, cl 10.10
Assess mitigations before significant risk increase.
Assessment before material changes — messaging No Tier (4A)
Sch 4A, cl 10.7
Assess mitigations before significant risk increase.
Online Safety Act 2021 (Cth) · Phase 2 Industry Codes of Practice · Head Terms + Schedules 4 & 4A · Diagram for reference only — refer to source documents for authoritative text.
© Zoe Huang. All rights reserved.