How AI chatbots can impact the insurance industry

0
94

Director from INZMO, a Berlin-based insurtech for the rental sector and a top 10 European insurtech driving change in digital insurance in 2023.

I believe future history books will likely name November 30, 2022, as the day everything changed. That was the day OpenAI launched ChatGPT in the wild.

It took people a few days to realize the leap forward from previous major language models (known as “LLMs”). The results people got made many realize that they could use this new technology to automate a wide variety of tasks.

Since then there has been a frantic battle to assess the possibilities. Just a few months after the release of ChatGPT (what I call “AC”), a questionnaire of 1,000 business leaders found by ResumeBuilder.com, 49% of respondents said they already use it. Almost all (93%) planned to expand their use of the technology. Another 30% of companies said they plan to start using it soon.

Contents

LLMs and the insurance industry

The insurance industry is a major employer of people. Many tasks in our industry required our incredible ability to solve problems quickly. We have to look for exactly the right information for a given situation and then communicate it to colleagues or customers in an understandable way.

Insurance giant Zurich has announced this already testing the technology “in areas such as claims and modeling,” said the Financial times (paywall). I think it’s fair to assume that most, if not all, of the other insurance companies are also looking at the technology. For example, my own company just launched a chatbot service to improve customer service.

What’s next for chat-enabled insurance?

If that’s where we can get to in a few months, what else can we expect?

LLMs can have a significant impact on the future of work, according to one Open AI paper. This also includes the insurance industry. The paper categorizes tasks according to their exposure to automation through LLMs ranging from no exposure (E0) to high exposure (E3).

LLMs can impact the insurance industry by automating some tasks, improving others and creating new opportunities for learning and innovation. Some of the insurance industry jobs that may be exposed to LLM-driven automation include:

• Writing and Editing Insurance Policies and Contracts: LLMs can help insurance professionals write and edit policies and contracts that are clear, accurate, and compliant with regulations. They can also generate templates and suggestions based on customer needs and preferences.

• Data Entry and Analysis: LLMs can help insurance professionals enter and analyze data from a variety of sources, such as forms, documents, and images. They can also extract relevant information, perform calculations and create reports and dashboards.

• Customer service and support: A chatbot can answer frequently asked questions, provide information, help solve problems and make personalized recommendations.

• Fraud detection and prevention: AI can help detect and prevent fraud by analyzing patterns, anomalies and behaviors in data. It is also possible to flag suspicious transactions, alert authorities and suggest actions to mitigate risks.

• Risk Assessment and Management: Natural language processing and machine learning can help insurance professionals assess and manage risk by understanding complex scenarios, evaluating probabilities and estimating outcomes. LLMs can also provide guidance and advice on reducing or transferring risk.

• Claims Processing and Settlement: LLMs can help verify the validity, correctness, and completeness of claims. They can also automate tasks such as document checking, damage assessment, payment calculation and communication with the claimants.

So, is everything going smoothly?

Hold on – this is not a one-way street and there are serious issues that need careful consideration.

For example, a February 2023 Ipsos survey of 1,109 US adults found that less than a third of respondents trust AI-generated search results. Insurers will need to convince and reassure customers about their use of LLMs.

To convince and reassure customers about AI, it is important that insurers are transparent about how they use the technology and what data they collect. Clearly explain how AI works and how it is used to make decisions. In addition, give customers the option to opt out of certain uses of their data or AI-based decisions. Insurers must also clearly inform customers about how their data is secured and what measures have been taken to prevent unauthorized access or misuse.

For leaders in the insurance industry, it is important to keep a close eye on the broader ethical picture to avoid the risk of consumer backlash. So when it comes to determining how and if they will use AI in the future, insurers need to identify the right use cases to ensure that the data used is accurate and unbiased. This requires being aware of the potential risks of AI and taking steps to mitigate those risks.

In broader ethical terms, there has been a lot of discussion about the explainability of AI, or rather the lack of explainability of AI. like a Insider intelligence article explained, “Generative AI is generally ill-suited to fully explaining its actions, making it unsuitable for making pricing decisions that need to be explained to internal stakeholders and regulators.”

Indeed, it is now clear that AI is far from perfect. It has limitations such as errors, biases, inability to understand context/nuances and ethical issues. Insider also pointed out that the “rapid rise” of AI means that regulations are currently lagging behind. It will catch up, but this will likely be piecemeal, with different approaches mandated in different national or state jurisdictions. This can lead to complex additional costs for insurance companies.

Despite these restrictions, and calls from some to slowing down the pace of AI development, I think AI chatbots are likely to play a bigger role in insurance in 2023. It is important for leaders in the insurance industry to keep ethics in mind. That might mean slowing down in the near term, but before November 30, 2022, could you really have imagined even getting that far?


gotechbusiness.com Business Council is the leading growth and networking organization for entrepreneurs and leaders. Am I eligible?