Current editionIndependent · Evidence led · Published in AustraliaIssue 005 · Your feed, your choice

The weekly paper
for curious people

See the connections.
Understand the consequences.

Free email via Substack →

Swipe to see more sections →

Issue 005 · The lead story

Australia promises a choice of feed. What would you actually control?

Lens · Edition of 12 September 2026 · Evidence checked 10 September 2026 · Six minute read

Proposed, with details still to come

The government proposes a choice over your default social media feed. The draft would let the minister specify the required tools later. It does not give you a new feed right today or settle every detail of how the choice would work.

Why Lens says this: promiseWhy Lens says this: delegated toolsWhy Lens says this: commencement
What people heard

The government says Australians would be able to opt out of personally recommended content in their default feed.

Is this a new right now, and what exactly would change?

Choose how to read this

Read, listen or follow the question.

The facts do not change. Each view uses the same published sources and leaves the same questions open.

AaReadGo straight to the best-supported answer.
Different viewsSee what each perspective notices—and may miss.

See through another Lens

Which view do you want to understand first?

Each view notices something useful. None is allowed to stand in for the complete evidence.

Reader

I want to choose what I see when I open the app.

What this view explainsWhy a lasting default matters as well as the existence of a separate tab.

What it may missChosen accounts can still supply misleading or incomplete information.

The promise concerns the feed you open.

On 8 September, the Australian Government announced that social media services would have to notify users and offer a choice for their default feed. People could choose personalised recommendations, or opt out of those recommendations and see friends and creators they follow.

The word default matters. A view you can visit is different from the view a service opens for you. The government is promising a choice about that starting point, rather than simply announcing that another tab should exist.

Why Lens says this: promiseWhy Lens says this: control distinction

The draft creates a power. The detailed tool comes later.

The proposal sits inside the draft Digital Duty of Care legislation. Its broader purpose is to place responsibilities on online services to address risks of harm. For this story, the important provision is proposed section 26.

Subsections 26(4) to (6) would allow the minister to require specified online services to provide specified user empowerment tools. A legislative instrument is a separate legal document used to set those requirements. The draft gives controls over recommended content as an example of what a tool could do.

This is a route to requiring feed choice. It is not a complete design for that choice. The text does not specify a button for each app, resolve the final ordering of posts or show how the choice would carry between devices. Those details will matter when assessing the final requirements and the services built to meet them.

Why Lens says this: delegated toolsWhy Lens says this: details open
The promise and the steps still needed
  1. AnnouncementA choice over the default feed is promised.
  2. Draft legislationThe proposed law permits tools to be specified.
  3. Later decisionsFinal law, detailed requirements and platform changes remain to come.

Lens explanatory diagram, based on the government announcement and exposure draft. The last stage is prospective.

When would it start?

There is no fixed calendar date in the records checked for this story. The department is seeking feedback on an exposure draft, with a stated deadline of noon on 22 September. The consultation page does not specify a time zone. The government says it intends to introduce legislation this year.

The draft commencement table puts the duty of care provisions in Schedule 2 after a period of 12 months beginning on Royal Assent. Assent is the formal approval of a bill after it passes Parliament. The draft may change before that happens, so it cannot support a promise that your app will change on a particular day.

Why Lens says this: draft statusWhy Lens says this: commencementWhy Lens says this: promise

Haven’t platforms offered a choice before?

Yes. Meta announced chronological Following and Favorites views for Instagram in March 2022. In July that year it described a Facebook Feeds tab alongside Home, which remained the opening tab. Its Feeds announcement also said ads would remain.

These are historical product descriptions, not a test of your current account. They establish a useful distinction: another view can exist without being your lasting opening preference. The proposed Australian requirement would need to be judged by the choice it actually secures.

Why Lens says this: existing controlsWhy Lens says this: control distinction

Does opting out mean every algorithm is gone?

That would go beyond what these records establish. The announcement concerns personally recommended content in the default feed. Selecting which accounts provide posts, choosing an order and choosing an opening view are separate decisions. Even newest first is a rule for ordering information.

Nor does this promise alone establish an end to advertising, data collection or recommendations elsewhere in an app. Those questions need the final requirements and the service’s actual settings. The practical companion lets you try the difference between account selection and ordering with fictional posts.

Why Lens says this: details openWhy Lens says this: control distinction

A useful choice is not a guaranteed cure.

A feed choice could help a reader spend more attention on accounts they deliberately selected. It could also reduce opportunities to discover unfamiliar accounts in that view. Which outcome a person values depends on why they opened the app.

A study of consenting Facebook and Instagram users during the 2020 US election found that chronological feeds changed exposure and activity. It did not find significant changes in measured political attitudes over three months. That result does not show that algorithms never matter, and it cannot predict the effect of this Australian proposal.

The next test is concrete: what can a reader select, does it remain selected, and what changes on the screen? Evidence about broader benefits comes after those questions, not before them.

Why Lens says this: control distinctionWhy Lens says this: research limitWhy Lens says this: outcomes open
How the public account compares with the draft

This is a comparison of three observed records, not proof of a complete chain of influence.

  1. The government announcement, 8 September, promises default feed choice.
  2. ABC reporting, 8 September, describes a proposed algorithm choice. Its framing preserves that this is a proposal.
  3. The draft, proposed section 26, supplies a power to specify tools. Lens adds the distinction between that power and the later details.

We have not established that readers generally misunderstood the proposal or that the reporting falsely presented it as law.

Why Lens says this

Australia promises a choice of feed. What would you actually control?

The government proposes a choice over your default social media feed. The draft would let the minister specify the required tools later. It does not give you a new feed right today or settle every detail of how the choice would work.

Last checked: 10 September 2026 · Draft policy with a bounded practical explanation

Direct record

The government announces a default choice; section 26 of the draft allows later specification of tools. Meta described separate feed controls in 2022.

Why Lens says this: promiseWhy Lens says this: delegated toolsWhy Lens says this: existing controls

Lens inference

The meaningful test is what a user can select, whether it remains selected and which content rules change. The promise alone cannot answer those questions.

Why Lens says this: details openWhy Lens says this: control distinction

Contrary evidence

The announcement does expressly promise a default choice, so it would be wrong to dismiss it as merely another tab. Existing controls also show this is not the first feed choice. The US study found changed exposure without significant changes in measured attitudes, cautioning against promises of an automatic social benefit.

Why Lens says this: promiseWhy Lens says this: existing controlsWhy Lens says this: research limit

What we checked

Seven records: government announcement, consultation page, exposure draft, ABC reporting, two historical Meta announcements and a research author abstract. These are not seven independent confirmations. No platform account, internal ranking system or Australian outcome dataset was inspected.

What remains unknown

  • What Parliament will enact and when it will receive assent
  • Which services and tools a ministerial instrument will specify
  • Whether the final alternative requires chronological ordering, and how advertising and other tabs are treated
  • How a choice persists across sessions and devices
  • How many people use the choice and what it changes for Australian readers

What would change the answer

  • A revised or introduced bill changes the proposed power or timing
  • A legislative instrument specifies feed tools and covered services
  • Platforms publish implementation details or observable controls
  • Research directly tests the resulting Australian feeds
How this explanation was prepared

AI assisted source comparison and copy review. Original public URLs and clause locations retained. Publisher reviewed and approved this issue on 12 September 2026.

A current proposal needs its promise, mechanism and unknowns kept separate; the companion compares practical choices without prescribing a preference.

  • Show the announcement, draft and remaining decisions separately. A policy promise is not the same record as a legal requirement.
  • An explicitly fictional feed example changes source selection and ordering independently. The illustration explains separate controls without claiming to reproduce a platform or forecast policy effects.
The sources and what each establishes

Government proposal

The government says users would be notified and offered a choice between personalised recommendations and followed friends and creators in their default feed.

Government announcement, 8 September 2026

My Feed, My Way ↗

Possible effects · We cannot say how likely

What this could change

It could give users more control over who supplies posts and which view opens first. Whether that changes discovery, accuracy, wellbeing or politics depends on final law, product design and measured use—not the announcement alone.

Documented action

The government announced a default-feed choice and the exposure draft proposed a power to specify user-empowerment tools by later legislative instrument.

What Lens thinks may follow

If final requirements secure a persistent and understandable setting, users could choose a followed-account starting view without relying on a secondary tab that resets.

Where the connection stopsThe draft creates a route to specifying tools; it does not settle the final interface, services, ordering rules, adoption or social outcomes.

What this depends on—and other possibilities

This depends on

  • The proposal becomes law and a later instrument specifies meaningful controls.
  • Platforms implement the setting so users can find it and their choice persists.

Other explanations

  • Platforms may provide formally compliant controls that few people use or understand.
  • Existing product controls may produce similar choices without the proposed law.
How different interpretations could affect what happens next

How people may respond

How the story itself could change what happens

How feed choice is interpreted could change whether people use the setting and what they expect it to control.

What the evidence does not showA choice mechanism can change ranking inputs; its effect on attention, wellbeing or political belief is not established.

One possible path

Feed choice is interpreted as meaningful control

Not enough evidence yet
  1. How it is told

    A persistent choice between ranked and alternative feeds is presented as a usable product decision.

  2. What people may take from it

    Users may pay more attention to settings and expect their choice to endure.

  3. Where attention could turn

    Awareness of ranking as a separate decision increases.

  4. What people may do

    Users, platforms and regulators: Use, expose and monitor the choice mechanism.

  5. What could change

    A stable setting can become part of ordinary platform control.

What we know has changedWe have not established that this possible change has happened.

What this does not showChoosing a feed does not control everything eligible to appear or prove a better outcome.

Why we are cautious
Why we cannot tell yet

This is the first time Lens has mapped this path. We have no later evidence showing whether it is happening more, less or about the same.

Signs that would support this path
  • Final instruments specify persistence, coverage, notice and the exact user controls.
Signs that would weaken it
  • The choice resets or is difficult to find.
  • The instrument covers few services or leaves the default unchanged.
This depends on
  • The proposal becomes law and a later instrument specifies meaningful controls.
  • Platforms implement the setting so users can find it and their choice persists.
One possible path

Feed choice is interpreted as complete control

Not enough evidence yet
  1. How it is told

    The language of taking back a feed is stronger than the bounded mechanics.

  2. What people may take from it

    Users may expect the setting to settle availability, order, exposure and effects at once.

  3. Where attention could turn

    Attention to remaining platform choices may fall.

  4. What people may do

    Users and platforms: Attribute later content experiences to the selected feed alone.

  5. What could change

    The control can feel ineffective if other ranking and availability decisions remain.

What we know has changedWe have not established that this possible change has happened.

What this does not showMixed outcomes do not prove the choice mechanism is meaningless.

Why we are cautious
Why we cannot tell yet

This is the first time Lens has mapped this path. We have no later evidence showing whether it is happening more, less or about the same.

Signs that would support this path
  • Independent studies measure adoption, exposure, activity and outcomes under the implemented Australian controls.
Signs that would weaken it
  • Robust evaluations find consistent effects across services and user groups.
  • Implementation differs materially from the studied feed designs.
This depends on
  • The proposal becomes law and a later instrument specifies meaningful controls.
  • Platforms implement the setting so users can find it and their choice persists.

What new evidence could change this view?

  • Passage or abandonment of the bill.
  • A final legislative instrument defining the controls.
  • Observable platform implementations.
  • Independent evidence about adoption and user outcomes.
Assessment 1 · We have not estimated how likely either path is.

Two ways this could develop

This depends on what happens next

Feed choice becomes a persistent default

If final rules cover major services and require a clear choice that remains selected

Then users could make deliberate account-selection and opening-view choices, while recommendations remain available elsewhere if the product permits them.

What to watch—and what would weaken it
  • Final instruments specify persistence, coverage, notice and the exact user controls.Enacted legislation, legislative instruments and observable platform settings.

Would weaken this: The choice resets or is difficult to find. The instrument covers few services or leaves the default unchanged.

Scope: Control over the default feed and recommended content, not all algorithms, ads or data collection. Horizon: The legislative, instrument-making and platform implementation cycle.

This depends on what happens next

Control improves while broader outcomes remain mixed

If people use the choice but followed and chronological views alter exposure in different ways for different purposes

Then some users could value control while evidence still fails to show a uniform improvement in accuracy, discovery, wellbeing or political attitudes.

What to watch—and what would weaken it
  • Independent studies measure adoption, exposure, activity and outcomes under the implemented Australian controls.Platform transparency reports, regulator research and peer-reviewed evaluations.

Would weaken this: Robust evaluations find consistent effects across services and user groups. Implementation differs materially from the studied feed designs.

Scope: Measured effects of implemented controls, not a universal ranking of feed types. Horizon: After implementation and sufficient independent evaluation time.

How do we know?Inspect the evidence and its limits

Evidence used in this assessment

Government announcement, 8 September 2026 · 8 September 2026My Feed, My Way government announcement

The government says users would be notified and offered a choice between personalised recommendations and followed friends and creators in their default feed.

Open evidence ↗
Department consultation page · 8 September 2026Digital Duty of Care exposure-draft consultation

The department released an exposure draft on 8 September and requests feedback by noon on 22 September 2026.

Open evidence ↗
Exposure draft, printed page 32 · 8 September 2026Online Safety Amendment (Digital Duty of Care) exposure draft

Proposed section 26(4) to (6) lets the minister specify services and user empowerment tools by legislative instrument, including controls over recommended content.

Open evidence ↗
Guess and colleagues, Science, 2023; author abstract · date unknownChronological-feed election experiment

A 2020 US election experiment changed exposure and activity when consenting Facebook and Instagram users received chronological feeds. The author abstract reports no significant change in measured political attitudes over three months.

Open evidence ↗

What could change this assessment?

  • Passage or abandonment of the bill.
  • A final legislative instrument defining the controls.
  • Observable platform implementations.
  • Independent evidence about adoption and user outcomes.

Where the evidence stops

Established hereA policy promise and exposure-draft route to later tools existed at the cutoff.

Not establishedAn enacted right, final interface or guaranteed social benefit.

Still unknownFinal coverage, product design, persistence, adoption and measurable outcomes.

Assessment as at 23 September 2026 · Evidence checked through 10 September 2026 · Revision 1

Free Lens updates

Understand the story. Get the next Lens.

Our free weekly email brings you the latest Lens reporting and what changed.

Get the free Lens email via Substack →
Delivered by Substack. Entering your email creates a private Substack account. No app or public profile needed. Unsubscribe anytime.
Open the signup page instead →