Regulatory sandboxes aren’t moving the needle for insurance

ASIC’s push to encourage insurance innovation is under review – and the early verdict from the market isn’t flattering.

In a submission to a Treasury review of the Enhanced Regulatory Sandbox (ERS), Insurtech Australia says the regulator has made a “genuine and sustained effort” to enable financial innovation – but that effort “has not made an impact on the insurance industry and insurtech ecosystem.”

No business providing insurance-specific services has successfully relied on the ERS, and claims handling isn’t even eligible – a major constraint when most meaningful innovation is happening in claims, underwriting and risk.

The timing matters. The independent chair of the ERS review and an ASIC senior executive are both speaking at InsurtechLIVE26 in Sydney on 18 February 2026. We’ll be there too – because this conversation goes to the heart of how insurance innovation really happens.

The blunt question is:

If regulatory sandboxes don’t move the needle, what does?

From where we sit – inside the operating stack of carriers, MGAs and brokers – the answer is clear:

The real sandbox isn’t at the regulator. It’s in your own systems, with the ability to rapidly prototype products, distribution and onboarding.

Why sandboxes struggle with real insurance innovation

The ERS lets firms test some financial services without a full AFSL, under strict limits. On paper, that sounds like a good way to de-risk experimentation. In practice, it’s structurally misaligned with how insurance innovation works.

1. Innovation needs scale, time and real-world mess

Insurtech Australia is direct: the ERS is “poorly suited to claims, underwriting and risk-management innovations, where learning requires scale, time and credible claim-loss experience.”

Most meaningful insurance innovation needs:

• Enough exposure to see real-world loss behaviour
• Enough time to watch performance over at least a renewal cycle
• Enough mess – claim leakage, behaviour change, operational friction – to know whether the idea actually works

The ERS caps exposure to $10,000 per retail client, which the submission calls “materially constraining testing of insurance propositions, where premiums are often low but risk exposure is high and can be long-tailed.” That might suit some fintech experiments; it’s poorly matched to parametric events, catastrophe covers or automated claims at meaningful scale.

2. It fixates on the product, not the enabling platform

The framework today is described as a “passive exemptions framework with weak transition pathways and an innovation test emphasising the underlying product rather than the enabling technology.”

But in most insurers, the product is not the main constraint. From our perspective, the hard part is:

• Can you configure and price it on your systems without a multi-month project?
• Can you embed it into existing and new distribution – broker portals, white-label partners, embedded flows?
• Can you onboard and service customers without breaking your operations?

That’s about the platform backbone you’re running on.

3. Complexity and weak scale-up pathways

The submission points to “complexity, fragmented oversight and limited pathways to scale” in the current regime, and notes that other countries embed sandboxes into licensing, supervision and market development more effectively.

Even if you manage to get something live in the ERS, the path from “test” to “scaled BAU” is slow and unclear. For a startup with limited capital – or a business unit with limited internal political capital – that’s often a deal-killer.

Where innovation actually happens: inside your operating stack

None of this means regulators shouldn’t try to support innovation. It does mean that waiting for the regulator’s sandbox to make innovation safe is a losing strategy.

For insurers, the practical sandbox looks very different:

• It runs on your core and embedded platforms – policy, billing, claims, portals – not off to the side.
• It supports rapid prototyping of products and distribution/onboarding flows, not just static wordings.
• It uses real or near-real data inside controlled cohorts, not only synthetic test sets.
• It’s bounded by clear governance and risk limits, not just legal exemptions.

In practice, your internal sandbox should let you:

1. Stand up product and rating variants quickly
   1. Clone an existing product
   2. Adjust rates, terms, triggers or conditions
   3. Apply them only to a defined test cohort or distribution channel
      
      
2. Prototype distribution and onboarding without a re-platform
   1. Spin up new broker or partner portals
   2. Embed flows into third-party journeys (e.g. lenders, real estate, platforms)
   3. Test different onboarding questions, UX and straight-through rules
      
      
3. Configure workflows, not hard-code them
   1. Add or change steps in quoting, binding, endorsements and claims
   2. Plug in new data sources or decision engines
   3. Route exceptions to humans without breaking the process
      
      
4. Track performance at the cohort/channel level
   1. Premium, loss ratios, conversion, lapse, NPS
   2. Time-to-quote, time-to-bind, time-to-settle
   3. Clear test vs control comparisons

This is exactly the backbone we’ve built at InsuredHQ: a platform that lets insurers, MGAs and brokers experiment in live or near-live environments, across both product design and go-to-market, without asking for a new core system project every time.

Innovation isn’t just “new insurance ideas”. It’s also how quickly you can try them with real customers, and retire what doesn’t work.

What insurers can do now – without waiting for ASIC

If you accept that the real sandbox is internal, the next step is to make it concrete. Here’s a pattern we see working in the field.

1. Define a tight use case, cohort and guardrails

Pick a specific innovation, for example:

• A parametric flood add-on for existing property policies
• A new SME package version sold only through a particular broker network
• A simplified digital onboarding journey for a niche segment

Then define:

• Cohort – which customers, which channels, which geographies
• Guardrails – max GWP, max exposure, customer eligibility rules
• Kill switches – pre-agreed thresholds on loss ratio, complaints, operational impact

2. Implement it as a platform configuration, not a project

On a configurable platform, your team should be able to:

• Clone and tweak products and rates
• Set up a new onboarding or distribution flow (e.g. branded portal, embedded widget)
• Route those customers into a distinct portfolio or book for tracking

If every small variant requires months of development and testing on an inflexible legacy core, you don’t have a sandbox; you have a change-control log.

3. Measure, decide, move on

Before going live, align on:

• Primary metric – e.g. combined ratio, quote-to-bind, onboarding completion, time-to-settle
• Decision checkpoints – e.g. after X policies, Y claims, or Z months
• Decision owners – who can say “scale”, “iterate” or “stop”

Then treat the initiative as an experiment: run it, review it, scale or shut it down. The value is not just in successes; it’s in how cheaply you can learn.

How regulators could help – in ways that map to reality

Insurtech Australia is not arguing to scrap the ERS, but to reform it: tiered pathways that distinguish startups from incumbents, risk-based limits for insurance tests, and better co-ordination across regulators so innovations like parametric products, AI-driven underwriting and automated claims can be tested without full obligations upfront.

A constructive direction from here would be frameworks that:

• Explicitly recognise internal operational sandboxes within insurers (configurable platforms plus governance)
• Align licensing, supervision and market development so that successful tests can scale more quickly
• Focus less on one-off “exempt” products and more on the platform capabilities that enable safe, repeatable experimentation

InsurtechLIVE26 in Sydney on 18 February will bring regulators, incumbents and insurtechs together to talk exactly about this: the role of technology in enabling insurance innovation. We’ll be in the room, because the bridge between policy intent and operational reality is where we work every day.

The bottom line

Regulatory sandboxes are useful signalling. They are not where real insurance innovation lives.

If you’re an insurer, MGA or intermediary in Australia or New Zealand, your real innovation leverage is your ability to:

• Rapidly prototype products and distribution/onboarding journeys
• Embed those experiments into your live operating stack, not just labs or pilots on the side
• Govern them with clear limits, visibility and accountability

That’s the sandbox that matters.

If you’re heading to InsurtechLIVE26 and want to compare notes on how organisations are already running low-risk experiments on live systems with InsuredHQ as the backbone, come find us in Sydney – or reach out beforehand and we’ll line up time.