How insurers can harness Artificial Intelligence

Once the realm of science fiction, Artificial Intelligence (AI) is quickly becoming part of our lives. In the business world, insurers can use AI to improve business models and the customer experience.

At its essence, AI is about embedding human intelligence into machines, enabling systems to learn, adapt and develop solutions to problems on their own. Given the promise it holds, global financing to AI start-ups has expanded an average of 62 percent annually since 2011, according to a recent estimate by UK technology provider Trustmarque.

In keeping with the Ubers, Apples and Amazons of the world – as well as the ever-burgeoning breed of innovative financial services and fintech firms before them – insurance companies are looking to shed their staid images and jump on the AI bandwagon. A number of established underwriters are already in the process of integrating AI into existing operations. For example, Swiss Re announced in late 2015 that it is working with IBM Watson to enable human advisors to spot emerging trends more quickly so that they can develop a targeted range of underwriting solutions and achieve more accurate risk pricing. Through a partnership with voice recognition company Nuance Communications, Manulife analyzes unique voice characteristics to create individual "voiceprints" for customers. When customers call in to access their accounts, there's no need for a conventional password and PIN – their voice is compared with a stored voiceprint. If there's a match, access is granted.

Meanwhile, new firms in the insurance sector have sprouted up, built on AI automation. enables customers to generate price quotes from a variety of insurers by simply texting a photo of their license plate. ZestFinance is pairing advanced analytics with non-traditional data sources to establish credit scores for people who may not have had credit in the past.

"""The success of an AI solution largely depends on the continuous learning it creates from every single business transaction or interaction it makes."

Insurance industry implications: From ‘so what,’ to ‘now what’

AI offers insurers multiple ways to grow their toplines. It can help them rapidly capitalize on promising new product lines, geographies and customer segments. For example, robo-advisors hold the potential to boost advisory effectiveness as well as efficiency dramatically. To improve their job performance, successful human advisors will use virtual assistants to tackle routine tasks so they can focus on higher-level challenges. And at a more profound level, insurers can transform their customers' experiences by being available 24/7, making it easier for customers to securely interact with them and offering leading edge, personalized solutions.

Embedding AI into insurance operations can also help reduce bottom lines. Most obviously, AI solutions enable organizations to reduce their manpower requirements and save overhead costs. AI-based systems can give insurers and customers alike faster access to information, quicker turnaround time and improved risk management, among other benefits. These improvements create operational efficiencies as well as service excellence. As a result, insurance company employees can focus on skilled tasks, building expertise and evolving the AI solutions.

Yet transforming a business with AI is not an overnight process. Insurance firms currently sit across a broad spectrum in terms of experience with new technology. Many of them must overcome a number of challenges to move to the next level.

  1. The first challenge for any insurer embarking on an AI project is to amass the right kind of information – and lots of it. For AI platforms to solve business problems, they need to be exposed to huge volumes of domain-specific information covering all possible business scenarios. The success of an AI solution largely depends on the continuous learning it creates from every single business transaction or interaction it makes.

  2. Challenge No. 2 is dealing with the risk of glitches – or inappropriate responses – when it comes to unleashing AI helpers. Even with good data inputs, advanced AI technology still struggles with accuracy and getting things right the first time. When a machine is incorrect, it can be wrong in a far more dramatic way, with more unpredictable outcomes, than a human could. Suppose a self-driving car is told to "get us to the airport as quickly as possible." Would the autonomous driving system increase its speed to 125 mph, putting pedestrians and other drivers at risk?

    In their initial phases, AI solutions will require that insurers make provisions for regular manual interventions and evaluate the existing state of their technology, building infrastructure to enable integration with AI solutions.

  3. The third key challenge is preparing various stakeholders for a new way of doing business. From an insurer standpoint, advisory, operations and contact center teams will undergo a major transition to align with AI implementations. Convincing employees to work with the technology, rather than spurn it, is no small feat when the technology threatens to take away their jobs. Insurers will need to carefully manage the process by redesigning tasks, jobs, management practices and performance goals, and setting employee expectations for how AI might change their existing roles.

    Customers are another important set of stakeholders. Customers will inevitably want to know if they can trust a machine with a long-term financial decision, and if bots and AI will be capable of dealing effectively with complaints and claims. It is incumbent upon insurers to engage with customers about new AI solutions and get ahead of their concerns with reasonable solutions and backup plans in case of AI glitches.

  4. Finally, to successfully implement AI, insurers must be ready to do battle with data privacy concerns. Robust AI solutions rely on learning from every interaction or transaction. However, most AI solutions are likely to reside on the cloud of a third-party technology provider, necessitating a transfer of sensitive information. Insurers will need to develop the appropriate information security and data privacy policies, procedures, methods and tools to protect data from breach or unintended use. They will also need to keep a close watch on legal and regulatory changes as they evolve to incorporate the risks associated with new AI technology.

Viewing AI through Multiple Lenses

The Foundationally Intelligent Insurer
These insurers currently rely heavily on traditional processes and legacy systems. They tend to resist major organizational change.

Our advice
Concentrate on more elementary techno-business issues that promise a better return on investment. Examples include Chat bots (starting with a "smart" configuration rather than NLP-enabled) and personal financial trackers and advisors (starting with a rules-based configuration) to target prospects and customers.

The Incrementally Intelligent Insurer
These insurers have invested heavily in IT solutions that enable "pull" marketing techniques rather than just "push" approaches. They know their customers' needs and behaviors and want to take this intelligence to the next level.

Our advice
Pilot basic AI solutions that can be built and tested in a short time span with a relatively low investment and risk of failure. Examples include virtual sales assistants that manage basic routine work (e-mails, meetings, lead search, etc.) and automated algorithms for needs analysis that can be deployed across all customer-facing channels, thus ushering in robo-advisors.

The Institutionally Intelligent Insurer
These businesses have advanced point-of-sale capabilities, straight-through processing functionality and a single view of the customer across all channels.

Our advice
These businesses have developed a foundation and are equipped to explore complex AI business solutions. Examples include claims transformation through intelligent prediction and adjudication and contact center modernization through voice recognition and interactions mining.

Looking ahead

To implement AI smartly, insurers should follow a series of steps before launching into a machine-learning project.

Establish a baseline

Assess the readiness of the organization, both from the technology and the cultural standpoints. What would it cost to support an AI solution? How do executives and business unit leaders feel about it? What are the longer-term implications for the operating models and product offerings?

Start small

Given the cultural and risk challenges facing the sector, insurers should start by developing a proof-of-concept model that can safely be tested and adapted in a risk-free environment. Since AI machines excel at routine tasks and their algorithms often learn over time, insurers should focus their early efforts on the processes or assessments that are widely understood. A second step would be to identify the right technology partner and AI solution to transform the identified use case from concept to reality.

Manage change

Because AI capabilities can potentially displace humans (or require talent upgrading), insurers need an effective and thoughtful HR strategy. Full communication and retraining of affected staff, as well as a focus on building new skill sets and training, will go a long way toward minimizing resistance and encouraging acceptance. This means insurers must focus on effective change management to ensure that impacted employees understand that the tools are deployed to help them do a better job, increase their productivity and value, and enhance customer satisfaction, which in turn will raise employee satisfaction and retention.

For insurers, AI truly has the potential to transform all areas of the business. While implementation may be challenging at first, the price of doing nothing is also increasingly high. With solid processes, thoughtful research and the right partners, the long-term gains from jumping on the AI bandwagon will almost certainly make up for any short-term pain it causes.

About the authors

Srinivasan Somasundaram is a Director within Cognizant Business Consulting's Insurance Practice. He has over 20 years' experience in the life, annuity and pension sectors. His experience includes business consulting, digital advisory, product development and managing business transformation programs.

Aritro Bhattacharya is a Manager within Cognizant Business Consulting's Insurance Practice. He has nearly 10 years of experience in the life, annuity and pension sectors, including consulting and delivery management, insurance IT product development, organizational change management, functional implementation and migration.

Divyaprakash Modi is a Consultant within Cognizant Business Consulting's Insurance Practice. He has six years of experience in the insurance industry across functions such as life insurance policy servicing, group business, life insurance IT consulting and business analysis.

About Cognizant

Cognizant (NASDAQ: CTSH) is a leading provider of information technology, consulting, and business process services, dedicated to helping the world's leading companies build stronger businesses. Headquartered in Teaneck, New Jersey (U.S.), Cognizant combines a passion for client satisfaction, technology innovation, deep industry and business process expertise, and a global, collaborative workforce that embodies the future of work. With over 100 development and delivery centers worldwide and approximately 255,800 employees as of September 30, 2016, Cognizant is a member of the NASDAQ-100, the S&P 500, the Forbes Global 2000, and the Fortune 500 and is ranked among the top performing and fastest growing companies in the world.

Discover more of our latest thinking.

This page was paid for by Cognizant. The editorial staff of CNBC had no role in the creation of this page.