ASIC's $678 Million Super Warning, ANZ's $7.59 Billion Profit and the AI Inflation Twist

Legally Insightful | Tuesday, 10 November 2026 | Australian law, finance, technology and consumer affairs

A superannuation sales funnel can move hundreds of millions of dollars. A bank can announce a multibillion-dollar profit and still worry about customers under pressure. And artificial intelligence might make things more expensive before it makes them cheaper.

Those are three of the more revealing findings from Monday, 9 November. Add Google's warning that Australia's proposed online-safety law could affect teenagers' access to Gmail, a new stage of open banking, and two international consumer-law developments, and there is plenty to unpack.

The important theme is not simply that regulators are busy. It is that the details matter: who is actually providing financial advice, whether a profitability measure has a fair comparison, when a new legal obligation commences, and whether an accusation has been proved or merely allowed to proceed in court.

Here are seven consequential stories, with their sources and practical implications.

1. ASIC's superannuation warning: $678 million moved through lead-generation models

What happened on 9 November: The Australian Securities and Investments Commission (ASIC) released Report 845, Lead generation advice models: Putting consumers and licensees at risk, together with a media release.

ASIC reviewed 15 Australian financial services (AFS) licensees whose financial-advice businesses used lead generators. Between January 2025 and June 2026, lead generators referred 7,639 new clients to financial advisers — 35% of new clients at the reviewed licensees. More than $678 million was switched or invested, about 90% from superannuation savings. The advice businesses received $9.8 million in initial fees, with commitments for another $4.5 million in ongoing fees.

The concern was not simply that financial advisers bought leads. ASIC found examples of unlicensed or unqualified people moving beyond marketing and introductions into activities such as fact-finding, assessing risk profiles and discussing advice recommendations. It also identified potentially misleading performance comparisons and pressure-driven sales tactics. Some licensees had weak supervision, including inadequate access to call recordings and poor monitoring of advisers' recommendation patterns.

Why it matters

Lead generation is not, by itself, prohibited. The legal line becomes dangerous when a marketer effectively performs personal-advice functions without appropriate authorisation, when a licensee cannot supervise its distribution chain, or when incentives undermine an adviser's obligation to act in the client's best interests.

For someone considering a super switch, a supposedly free health check, an urgent phone call or an advertisement claiming their fund is underperforming should not be mistaken for independent financial advice. Ask who is authorised, how the person is paid and whether comparisons use like-for-like investments, risks and fees. ASIC's Moneysmart guidance explains how to check an adviser and assess approaches about superannuation.

For AFS licensees, practical controls include approved scripts, access to recordings, testing referral sources, review of advice patterns, capacity controls and clear contractual boundaries. ASIC's findings concern the 15 reviewed licensees, not a finding that every advice business or every lead generator has broken the law. Proposed government reforms are separate from the obligations ASIC can already enforce.

Bottom line: Where a sales funnel influences a retirement decision, the responsibility cannot disappear between the marketer and the licensed adviser.

2. ANZ's $7.59 billion cash profit comes with a warning about credit pressure

What happened on 9 November: ANZ released its full-year 2026 results for the year ended 30 September.

The bank reported statutory profit of $7.270 billion and cash profit of $7.591 billion. Cash profit rose 31% against the prior year on an unadjusted comparison, but only 10% excluding significant items recorded in FY2025. Those two figures are both correct; they answer different comparison questions.

The board declared a final dividend of 83 cents per share, 80% franked. The full-year dividend totals 166 cents per share. ANZ reported a Common Equity Tier 1 (CET1) capital ratio of 11.99% at 30 September 2026.

But the result was not just a celebration of earnings. ANZ's full-year credit impairment charge increased to $582 million, comprising $360 million in individual provisions and $222 million in collective provisions. The latter was $108 million higher than in FY2025, reflecting growth in the portfolio and emerging macroeconomic risk. ANZ said most Australian retail customers remained ahead of scheduled repayments, while financial pressure became more evident for some in the second half.

Why it matters

A bank's profit and customers' financial resilience are related, but they are not identical. Higher profits do not establish that every mortgage holder is comfortable. Equally, an increase in impairment charges does not mean that all those customers have defaulted: provisions recognise expected credit losses, including portfolio-level risks.

For investors, it is worth looking beyond a single headline growth rate to the previous year's one-off items, credit quality, capital ratios and dividend coverage. For consumers, the issue is how persistent higher rates and living costs affect repayments. For businesses, it is whether a lender's risk appetite or pricing changes as the credit cycle evolves.

Bottom line: The striking part of ANZ's result is the coexistence of strong reported earnings and increasingly visible pockets of borrower stress.

3. RBA: artificial intelligence could raise inflation before delivering productivity gains

What happened on 9 November: Reserve Bank of Australia Assistant Governor (Economic) Sarah Hunter delivered a speech at the UBS Australasia Conference titled What Happens to the Economy as 'Intelligence' Gets Cheaper?.

Her message challenges the popular idea that AI must immediately make the economy faster, cheaper and more productive. The RBA sees a substantial investment boom in data centres and supporting infrastructure, but says evidence of a material lift in economy-wide productivity is still limited. Its estimates of realised Australian data-centre investment have increased tenfold over three years, albeit from a low starting point.

The near-term effect can be inflationary: building data centres, generating electricity, constructing networks and acquiring specialised equipment all create demand before efficiency gains have fully arrived. Over time, more capacity and better productivity could ease those pressures. Hunter also said there was little evidence of large effects on aggregate employment so far, although some hiring indicators for younger workers in AI-exposed occupations deserve watching. She explicitly cautioned that it is too early to attribute those modest differences to AI.

Why it matters

Technology can raise the economy's potential output, but implementing it requires capital, skills, electricity and time. A company may spend heavily on servers and redesign workflows before seeing a productivity dividend. That matters to inflation forecasts, infrastructure planning, the power grid and investment valuations.

The speech is an economic assessment, not a fresh RBA interest-rate decision or proof that AI has caused a particular number of job losses. It builds on, rather than repeats, earlier ABS data about reported effects on vacancies. It also shows why boards should distinguish an AI business case from the enthusiasm surrounding AI infrastructure.

Bottom line: The productivity upside may be real, but the construction bill — and its inflationary effects — arrives first.

4. Could Australia's online-safety bill mean teenagers lose Gmail? Google says it is a risk

What happened on 9 November: At an Australian Senate hearing, Google warned that the proposed Digital Duty of Care legislation could have unintended consequences for ordinary online services, including email, messaging and private file storage. Reuters reported the testimony on 9 November.

Google's Australian and New Zealand policy representative argued that age-assurance requirements tied to the risk of children seeing pornography could lead to under-18s being blocked from services such as Gmail, or encourage intrusive checks of private communications. Meta, TikTok and Snap also raised concerns about breadth and regulatory discretion. The government argues that technology providers need strong obligations to minimise serious online harm.

The government's explanation when it introduced the bill on 20 October describes minimum online-safety standards and responsibilities covering a wide range of digital services. Reuters reported that the bill had passed the House of Representatives and still required Senate approval; it also reported potential penalties of up to A$109 million for non-compliance.

Why it matters

There is a genuine legislative tension. Children need meaningful protection against dangerous material, while privacy and ordinary communication services should not be swept up by vague or disproportionate compliance measures.

Crucially, Google was warning about a potential interpretation or consequence. It was not announcing that Gmail had stopped working for minors, nor that the proposed legislation expressly mandated a Gmail ban. Parliament, regulators and providers still need to resolve scope, thresholds, proportionality and implementation.

For product teams, the lesson is that broad duties require operationally precise definitions: which services are covered, what risks activate age checks, how privacy is protected, and what evidence demonstrates reasonable compliance.

Bottom line: Safer online services and private communication are not mutually exclusive aims — but drafting needs to make them compatible.

5. Open banking expands: some non-bank lenders must start sharing customer data

What changed on 9 November: The consumer-data-sharing phase of Australia's Consumer Data Right (CDR) began for the first group of non-bank lenders. This is a scheduled legal commencement, not a newly issued ACCC media release.

The ACCC's original announcement of 13 July 2026 said non-bank lenders had begun phased sharing of product data on that date and that consumer data sharing would start from 9 November 2026, depending on the provider category. The ACCC's CDR guidance confirms the staged rollout; later obligations apply to other classes of provider.

CDR allows eligible customers to authorise the secure transfer of relevant information held by a participating lender to an accredited recipient. That can make it easier to compare loans, demonstrate financial position and seek alternative offers without manually collating every statement. It is opt-in, not general permission for lenders to sell customer data.

Why it matters

Until now, comparing non-bank loan products and sharing customer financial histories could be cumbersome. The change is particularly relevant to borrowers using specialist mortgage providers, vehicle finance or personal loans. For non-bank lenders, compliance involves data accuracy, technical readiness, authorisation controls, consent management and security.

Not every non-bank lender becomes subject to identical consumer-data obligations on the same date. Consumers should check whether their provider and intended recipient participate before assuming data-sharing functions are available.

Bottom line: A scheduled compliance date can matter as much as a new announcement — and the practical value is stronger consumer control over financial data.

6. Europe considers a digital-fairness law targeting 'dark patterns' and addictive design

What was reported on 9 November: Reuters obtained details of a draft European Commission Digital Fairness Act. The reported draft would expose companies to penalties of up to 6% of annual worldwide turnover for certain unfair digital practices.

The proposed measures include restrictions on manipulative website and app design — often called dark patterns — and personalised advertising directed at children and vulnerable consumers. The reported draft would also require greater transparency about stock available at advertised dynamic prices and allow users to disable certain addictive product features.

Why it matters

Many online purchasing decisions are shaped by interface design. A countdown timer, repeated prompt, confusing unsubscribe route or poorly disclosed variable price can change consumer behaviour even when the underlying product is lawful. The regulatory direction is therefore towards examining how people are persuaded, not just what a seller says.

Status matters: As at this edition, Reuters was reporting on a draft seen ahead of its planned presentation on Wednesday, 11 November. It was not an enacted EU law, and its final content would depend on negotiations between EU institutions. Australian businesses operating in Europe should track developments rather than treat the reported draft as already binding. Australia has its own unfair-trading reforms with a different legal timetable.

Bottom line: Regulators increasingly see website design as a consumer-protection issue, not simply a marketing choice.

7. A US judge allows key claims in a lawsuit over Claude's subscription advertising

What was newly reported on 9 November: Reuters reported a California federal court ruling in a proposed class action against Anthropic over its Claude Max subscription marketing. The ruling itself was issued on Friday, 6 November, and was reported by Reuters on Monday, 9 November.

Subscribers allege that advertising of '5x or 20x more usage' led them to expect corresponding increases over a longer period, while the higher-priced plans had usage caps that did not scale in that way. Anthropic disputes that interpretation, saying the comparison referred to per-session usage and pointing to further disclosures.

US Magistrate Judge Alex Tse allowed certain claims, including deceptive-advertising, negligent-misrepresentation and contract claims, to proceed. Other claims were dismissed, and the plaintiffs were allowed to amend their pleading.

Why it matters

The legal question is about what a reasonable customer could understand from a prominent comparison and whether limitations were explained clearly enough at the purchasing decision. It has obvious relevance to software subscriptions, cloud-computing plans and AI products where usage limits can be complicated.

A motion-to-dismiss ruling is not a finding that Anthropic misled anyone. It means specified allegations were sufficiently arguable to continue to later stages of litigation. The case is Kahn and Smith v Anthropic PBC, US District Court, Northern District of California, No. 3:26-cv-05763-AGT.

For any business advertising 'five times more', the lesson is simple: specify five times what, over which period, and subject to which caps — preferably where the customer can actually see the qualification.

Bottom line: A big headline claim can create risk when the small print defines a materially narrower product.

---

The bigger picture: disclosure is not the same as understanding

Today's stories look unrelated: superannuation, bank earnings, inflation, online safety, data portability and software subscriptions. But each turns on the difference between a technical description and a consequence that people can actually understand.

A lead generator may call itself a marketer while effectively steering a financial decision. A bank may report 31% profit growth while its underlying comparable increase is 10%. An AI model may be advertised as improving productivity while its supporting infrastructure adds to demand and costs. A law intended to protect children may be drafted broadly enough that ordinary communication services fear collateral consequences. A five-times-usage claim may depend on a time period buried away from the sales page.

Good law, good governance and good journalism all ask the same follow-up: what does the claim mean in practice?

Frequently asked questions

Is paying for financial-advice leads illegal in Australia?

Not necessarily. The risk arises when unlicensed people perform regulated advice activities, when claims mislead consumers, or when an AFS licensee fails to meet its supervision and best-interests obligations. ASIC's 9 November report examined particular models used by 15 licensees.

Did ANZ's profits really rise 31%?

ANZ reported a 31% increase in cash profit against its prior-year reported figure. Excluding significant FY2025 items, the increase was 10%. Its statutory profit and cash profit are different measures.

Has the RBA decided AI causes inflation?

No. Sarah Hunter described a likely near-term upward pressure from AI investment and supply constraints, with potentially offsetting longer-term productivity effects. It was an economic speech, not a new rate decision.

Are Australian teenagers about to be locked out of Gmail?

Google warned about that possibility under one interpretation of the proposed Digital Duty of Care legislation. No Gmail ban was announced, and the bill still required Senate consideration at the time of the hearing.

Can every non-bank borrower share data through CDR from 9 November?

No. The obligations are phased and depend on provider category and readiness. Consumer-authorised data sharing commenced for the initial group on 9 November, while others have later dates.

Is the EU's proposed 6% penalty already law?

No. Reuters described a draft Digital Fairness Act expected to be presented on 11 November. Legislative negotiations would still be needed.

Has a court found Anthropic liable for false advertising?

No. The court allowed certain pleaded claims to proceed. Liability has not been established by that procedural ruling.

---

Source and timing note: This edition covers the period 9 November 2026, 5:00 am to 10 November 2026, 5:00 am Brisbane time (AEST). Five Australian items concern a publication, speech, result, hearing or legal commencement on 9 November. The EU draft was reported on 9 November. The Anthropic procedural order dates to 6 November, but the identified new news report was published on 9 November. The ACCC's CDR announcement was originally published on 13 July, and is included solely because the previously announced consumer-data obligation commenced on 9 November. Earlier topics from the previous daily editions have not been repeated as standalone stories.

General-information disclaimer: Legally Insightful provides general news and analysis only, not legal, financial, taxation, investment or personal advice. Allegations and pending proceedings are not findings of liability; forecasts and draft legislation are not settled outcomes. Requirements depend on applicable law and individual facts. Seek suitably qualified professional advice before acting on information in this article.

Legally Insightful — Law, money and technology, explained.