InsurTech4Good.com Weekly Newsletter – #35, 2026

Welcome to all new readers. There are now around 4,500 of you!

This is a super-packed issue. A lot has been happening recently in finance and insurance innovation, from both a regulatory and business perspective, and I hope this helps you navigate the space a bit.

This issue is about agentic AI going enterprise, what supervisors are learning (and testing) in practice, and how policy strategy documents from Hong Kong to Europe are increasingly framing innovation as the plumbing that enables safer, faster change, not just a “nice to have”.

Hope you enjoy the read. Please let me know if you need help, or if you have suggestions on how to further improve the newsletter.

Andres

Subscribe to my newsletter here
See how I can help you here
Follow me on LinkedIn
andres@insurtech4good.com

OpenAI Frontier and insurance

OpenAI just launched “Frontier” and State Farm is a launch partner.

This matters because Frontier isn’t another chatbot. 

It’s positioned as an enterprise platform to build, deploy, and manage AI agents that support real workflows, with shared context, onboarding and learning, and clear permissions and guardrails.

State Farm’s message is telling: the goal is to strengthen the capabilities of agents and employees so they can deliver best-in-class customer service.

Read more here. 

Agentic AI in insurance

A quick note on Agentic AI in insurance.

When an operational process is fundamentally inefficient, deploying AI on top of it rarely delivers sustainable value.

First, redesign the process. Then apply automation and AI where they genuinely improve outcomes.

Boring, I know. But you’ll thank me later.

Read more here. 

Hong Kong Fintech Promotion Blueprint

Countries across the globe are exploring how to tackle financial innovation.

It’s not merely a “nice to have”.

It can help address structural change in finance and insurance, strengthening overall financial health.

And that’s not a “nice to have” either.

Done well, innovation can bring more people into financial services and support wider economic growth, which in turn benefits society as a whole.

Some will argue that strategy documents are irrelevant.

But if a country is aligned on where it wants to go, its risk tolerance, and how roles are divided across the ecosystem (who does what), the chances of success are always higher.

A recent example from Hong Kong is the Hong Kong Monetary Authority’s FinTech Promotion Blueprint.

It focuses on AI, DLT, high performance computing, data excellence, and cyber resilience, backed by practical initiatives such as a Fintech Cybersecurity Baseline, a New Risk Data Strategy, and a Quantum Preparedness Index.

Read more here. 

9 key findings from EIOPA Generative AI in Insurance Survey

1. Gen AI adoption is widespread and growing rapidly; 65% of insurance undertakings are already actively using Gen AI systems, but the majority of use cases are at a proof-of-concept stage, highlighting its growth potential.

2. Efficiency is the main driver; insurers are primarily adopting Gen AI systems to enhance the efficiency of internal processes and reduce costs, followed by enhancing customer interactions and improving decision-making.

3. Privacy, regulation, and talent are the key barriers; data privacy and security concerns, regulatory compliance (such as the GDPR), and a lack of skilled talent are the most significant reported challenges to implementation.

4. Focus is primarily on back-office operations; 64% of the reported use cases are for internal back-office applications (e.g., productivity tools, coding assistants, agent support) compared to 36% for customer-facing applications.

5. Use cases span across the value chain; the most active areas for Gen AI use are customer service, claims management, and sales and distribution. Fraud detection is the area with the highest planned future adoption.

6. Human oversight remains dominant; current adoption is dominated by "Assisted" models requiring human oversight. A shift is expected towards "Semi-Autonomous" and “Agentic AI” systems in the medium term.

7. "Hallucinations" are the top-cited risk; insurers identify inaccurate outputs as the main risk of Gen AI systems, followed by cybersecurity risks, data protection, and lack of explainability.

8. Need to adapt existing governance and risk management frameworks; 49% of undertakings have developed a dedicated AI policy (a twofold increase from 2023), with Gen AI systems requiring a greater focus on the model inference stage (e.g. prompt engineering and outcomes monitoring).

9. Reliance on third-party providers is high; the dominant ä strategy is purchasing off-the-shelf solutions or building on third-party models, making vendor risk management crucial. The majority of insurers view the AI Act's provisions about Gen AI systems as useful for ensuring provider reliability.

Although the data is from summer 2025 and developments have moved quickly since then, it remains an excellent read, particularly the sections on governance and risk management.

Read more here. 

AI Live Testing in finance

The pace of technological and innovation-driven change in financial and insurance services increasingly requires bold thinking from policymakers and regulators. Traditional supervisory tools may not always be well suited to these developments.

One example is the FCA’s AI Live Testing initiative, delivered as part of the AI Lab. It examines the challenges firms face before and during live AI deployment in UK financial markets.

The programme is designed to foster industry–regulator collaboration, build a shared understanding of the risks and practical challenges associated with AI design and deployment, and explore potential evaluation methods.

It also considers frameworks that could support safe and responsible AI across the full lifecycle from input quality controls to outcomes measurement.

AI Live Testing also gives participating firms access to relevant AI technical expertise alongside regulatory expertise, while helping identify emerging regulatory challenges that may arise in practice.

While forms of sandboxing are not new, what stands out here is the explicit emphasis on outcomes: testing an AI system against its intended purpose is recognised as good business practice, but it also helps deepen supervisory understanding of what “safe and responsible AI” should mean in operational terms.

Read more here. 

The World’s First Fully Open, Accelerated Set of Models and Tools for AI Weather

NVIDIA has just announced a new Earth-2 family of open models and tools designed to make AI weather forecasting production-ready and deployable on your own infrastructure, from ingesting observations to 15-day global forecasts and local storm nowcasts.

Why it matters, especially for risk-heavy sectors: faster, cheaper, and more customizable forecasting could enable better decisions across energy, agriculture, and public health, and strengthen financial and insurance risk assessment. 

The post references work involving AXA and S&P Global Energy.

Read more here. 

Review into the long-term impact of AI on UK retail financial services

The UK financial conduct regulator, the FCA, is reviewing how AI may reshape retail financial services for consumers, firms, markets, and regulators by 2030 and beyond. 

Read more here. 

AI-first companies win the future?

Not “AI-first companies win the future,” but consumer-first financial and insurance companies win the future.

AI is an enabler, not the strategy. AI-first can optimise the wrong thing faster; consumer-first defines the right thing to optimise.

Where we really are in practice is, to me, harder to pin down.

I see three camps: one is largely ignoring this and continuing with business as usual; another is heavy on hype but light on tangible results; and then there’s a middle group quietly building real solutions. Agentic workflows, agent networks, and measurable outcomes. Without talking too much about it.

What I think we’ll see is the technology gap widening even further.

From a public policy perspective, that’s a concern: this fragmented, two-speed development can ultimately become a supervisory, regulatory, and potentially even a financial stability issue.

Which raises the question of what role regulators should play in innovation facilitation. Barrier removal, enabling conditions, and ensuring the laggards don’t become a systemic weak point.

Read more here. 

Insurance Top Trends 2026

Capgemini has published Insurance Top Trends 2026. 

My personal number one recommendation for financial and insurance providers remains simple: talk more to your consumers.

Understand their needs, wishes, concerns, and pain points and act on that. It gives you a better roadmap than any generic checklist.

One number from the Capgemini piece that stuck with me: ~60% of customers say they’re willing to share personal data for more tailored coverage.

Worth keeping in mind as our open finance/FIDA discussions occasionally enter “winter sleep”.

Read more here. 

International insurance supervision priorities

International insurance supervision priorities this year: digital innovation + operational resilience.

If you work in insurance, InsurTech, or supervision, these IAIS priorities are worth tracking:

1) IAIS FinTech Forum (ongoing focus)

AI: emerging AI trends in insurance + finalisation of an AI question bank to support supervisors’ engagement with insurers on AI use cases.

SupTech: governance processes and the use of AI in SupTech, including a members-only report on agentic AI to help identify prudential and conduct risks.

2) Operational resilience (early 2026)

Following two public consultations, the IAIS is expected to publish an Application Paper on operational resilience objectives and a toolkit to support authorities’ oversight of operational resilience in the insurance sector.

3) Third-party risk (next step)

Building on the above, the IAIS will develop a members-only analytical report on emerging third-party trends and practices in the insurance sector.

Beyond that, I’m also closely following fair value for consumers and inclusive insurance because meaningful innovation often plays a key role in solving both.

Read more here. 

EIOPA Annual Work Programme 2026 and innovation activities 

16 innovation-related activities from the just published EU insurance watchdog EIOPA Annual Work Programme 2026 you should know:

1. Contribute to the implementation of the European Single Access Point (ESAP).

2. Assess how SupTech tools can enable the extraction and use of data from ESAP.

3. Develop new SupTech and use of AI Program in EIOPA and with NCAs.

4. Support development of innovation, use of AI and SupTech at NCAs and EIOPA, including through shared tools and infrastructure.

5. Further develop sharing of practical SupTech solutions, including coding, and explore infrastructure for this.

6. Implement, depending on the outcome of the testing, SupTech tools to detect possible greenwashing.

7. Promote the digital transformation and use of satellite data to supervise catastrophic events.

8. Pursue thematic work on digital touch points between consumers and insurance and pensions sector.

9. Monitor possible regulatory barriers impacting innovation.

10. Monitor digital market innovations.

11. Further improve methodological tools for the assessment and reporting of risks in the insurance and pensions sectors based on different econometric techniques exploring the development of Artificial Intelligence and Machine Learning Techniques.

12. Support and monitor the implementation and supervision of the AI Act.

13. Further clarify the regulatory framework on the use of AI by insurance market and support NCAs in its supervision.

14. Continue work foreseen under potential mandates as part of the Regulation on a Framework for FIDA (where mandates are agreed in the course of 2025).

15. Monitor distributed ledger technology (DLT)/Blockchain, focusing on Decentralised Finance developments in insurance and link to MiCA.

16. Monitor the use of crypto assets in the insurance and pension sectors and support the implementation of MiCA and issuing opinions at the request of NCAs as relevant.

Overall, these activities sit under EIOPA’s push to leverage SupTech and digital innovation to make regulation and supervision simpler, bolder, and faster.

They focus on strengthening risk detection and data-driven decision-making through technology.

They also aim to speed up digital transformation at EIOPA and national supervisors through shared tools, common standards, and capacity-building, while assessing the risks and opportunities of the industry’s digital shift.

Read more here. 

Supervision of artificial intelligence in finance

An excellent deep dive into how financial supervisors are approaching AI in finance and insurance, and the real-world challenges of applying AI regulation in practice.

The report strikes a thoughtful balance between enabling responsible AI adoption and preserving market stability, integrity, and consumer protection.

Read more here. 

Financial inclusion and open finance

A three-billion-person challenge (and opportunity!) for financial-sector policymakers and financial service providers.

Around three billion adult account holders are not fully engaged with the formal financial sector.

An urgent global challenge and a unique opportunity to advance economic inclusion at scale and accelerate progress toward the Sustainable Development Goals (SDGs).

Innovation and digital tools such as open finance and open insurance, mobile access, and AI can help unlock inclusion, boost economic growth, and expand opportunity, but only with strong policy frameworks, responsible oversight, and consumer protections in place.

And this is what I’m so excited about in digital finance and insurance. Whether from the regulatory, policy, or technology side.

The potential to deliver impact at this scale. Impact that counts.

Now back to work.

Read more here. 

EU Data Act

The Data Act establishes a horizontal set of rules on data access and use that respect the protection of fundamental rights and deliver wide-ranging benefits for the European economy and society.

It increases data availability, particularly of industrial data, and encourages data-driven innovation, while ensuring fairness in the allocation of data value among all actors in the data economy.

The Commission has now published updated Q&As.

Read more here. 

Open Finance in Korea and Georgia 

As I’ve said, the future of finance and insurance is open and successful are those who realize it today! And while progress on Financial Data Access (FiDA) discussions in the EU has slowed somewhat, some countries are already considering how to implement FIDA-type regulation in practice.

This study is a comparative analysis conducted to explore how Korea’s MyData framework can be incorporated into Georgia’s open banking and open finance development strategy. 

Korea has established a user-centric data ecosystem that spans finance, telecommunications, healthcare, and the public sector. This ecosystem is supported by a world-leading Data Re-use Index, a mature MyData infrastructure, a stringent licensing data-protection regime, and extensive API standardization. 

In contrast, Georgia remains in the early stages of PSD2-based open banking and will need to transition toward PSD3 and FiDA framework. 

Drawing on Korea’s experience, this study presents policy recommendations for Georgia’s transition toward a data-driven economy, including regulatory design, API standardization, governance structures, the scope of data provision, fee and settlement mechanisms, licensing requirements, and consumer-protection frameworks.

Read more here. 

EIOPA's strategy towards 2030 and innovation

The European Insurance and Occupational Pensions Authority (EIOPA) has published its new strategy.

Looking ahead to 2030, EIOPA’s top priorities will be to strengthen the single market, improve society’s resilience to risks, and enhance regulatory and supervisory effectiveness.

The strategy refers to leveraging SupTech and digital innovation.

More concretely, EIOPA aims to enhance supervisory effectiveness and operational efficiency by:

1. Leveraging supervisory technology (SupTech) to strengthen risk detection, monitoring, and data-driven decision-making.

2. Accelerating the digital transformation of EIOPA and national competent authorities (NCAs) through shared tools, capacity-building initiatives, and common digital standards.

3. Supporting the digital transformation of the industry by assessing related risks and opportunities.

I will, of course, follow especially closely how this part is implemented in practice.

I also believe it will be difficult, if not impossible, to address any of the major risks and challenges ahead (as the rest of the strategy rightly highlights) without truly embracing innovation, putting it into practice, and clearly supporting market participants, whether we call it facilitation or something else.

Read more here. 

Top 10 global risks for 2026

Both the short- and long-term outlooks of leaders and risk experts point to “turbulent” to “stormy” times ahead with geopolitics, mis/disinformation, polarization and cyber in the near term and climate and nature-related risks dominating the longer view.

For the insurance industry, this is the dual role in practice: these risks increase volatility and loss severity, but they also create opportunities to innovate through better risk insights, stronger prevention and resilience services, and new protection products that match how risks are evolving.

Read more here. 

Global Cybersecurity Outlook 2026

I’m still refining my areas of focus for the coming year, but one topic is definitely rising: cyber-enabled fraud.

Driven by generative AI, it’s becoming more sophisticated, easier to execute, and cheaper to scale.

From a public policy debate perspective, we’re probably also seeing more discussion about the responsibilities of different players across the value chain (telcos etc).

On the payments side in particular, I think we will see growing debate about whether we should slow down payment initiation to reduce fraud risk, and where to draw the line between prevention and friction.

That balance is genuinely hard to get right. Personally, I don’t want my payment flows slowed down rather vice versa, but it’s also easy to say that if you haven’t been directly affected by payment fraud.

Either way, I think we’ll continue to see, and should see, more serious debate around this.

At the same time, the very same AI and technology enabling fraud can also support solutions, from improved prevention and detection to more tailored tooling from specialist vendors, often in collaboration with telcos and other data-rich players.

From an insurance perspective, there still seems to be a meaningful protection gap in how cyber risks are addressed. But there’s also likely more scope for insurers to shift further upstream into prevention: getting involved earlier to help organisations assess vulnerable points, evaluate controls and tech stacks, and reduce exposure before losses occur.

This is admittedly a rough reflection, prompted by a report I just read. I don't know. I’m still thinking this through, but the direction feels right.

Read more here. 

European Supervisory Authorities and UK financial regulators sign Memorandum of Understanding on oversight of critical ICT third-party service providers under DORA

European Supervisory Authorities and UK financial regulators signed Memorandum of Understanding on oversight of critical ICT third-party service providers under DORA.

This is an important step for the financial and insurance sectors, as these issues are inherently global and increasingly cross-border and cross-sectoral.

Read more here.