Not for long: as one report by McKinsey analysts notes, the notion that insurance is a low-engagement, disintermediated category in which customer relationships can be delegated to agents and brokers is increasingly obsolete: digital disruption is arriving hard and fast.

As a result, insurance now ranks in the top quartile of sectors likely to be disrupted in the next few years, one expert, Jean-François Gasc, says. Digital disruption may have not arrived at the gates, but it is throwing ladders against the walls and climbing fast.

Gasc notes: “We calculate that the approaching shake-down could cost some as much as 40 percent of their traditional risk-protection revenues. This could happen within five years. Insurers need to react quickly to protect themselves. Those that move swiftly can gain the upper-hand over their more sluggish competitors. They’ll not only be able to preserve many of their revenue streams. They could also tap exciting new business opportunities”.

Insurgent Insurtechs Pose a Threat and Teach Lessons

Changing market dynamics and the growing impact of insurance technology startups called insurtechs are, meanwhile, forcing traditional insurers to move from a product focus to a customer-centric philosophy. As insurers examine new business, operating and organisational models, industry leaders are embracing insurtechs rather than competing against them.

As McKinsey puts “digital technology and the data and analysis it makes available give insurers the chance to know their customers better. That means they can price and underwrite more accurately, and better identify fraudulent claims. They can also offer clients more tailored products—and they can offer them in a more timely manner.”

IBM’s Stefan Riedel, Vice President, Insurance and Insurance Solutions Europe agrees: “The proliferation of usage-based services, such as hourly car insurance, mobile microinsurance and hotel-rental coverage also reflects the shift away from traditional risk-calculating insurance product lines and organizations toward richer and more personalized options. As instigators of many of these changes, insurtechs have become a crucial source of innovation for the global insurance industry. Insurers that don’t embrace the power of insurtechs may find themselves threatened not only by insurtechs themselves, but also from entrepreneurial insurers that employ insurtech services.”

Yet as a recent whitepaper from IBM that Riedel co-authored warns, “inflexible existing systems hobble the ability of many insurers to move forward.” To succeed, IBM notes, insurance companies need agile platforms, technologies and tools to move successfully into the future.”

Taking insurance from a product orientation to a client-centric model can be facilitated by a transition from traditional to platform business models.

Because platforms enable connections between producers and consumers directly, they enable organizations to reduce constraints to growth, and generate higher profits. By providing innovative services to customers through new channels, new digital offerings, in turn, also can provide insurers with new insights into the customers they serve.

According to the 2018 IBM Institute for Business Value C-suite Study, organizations across every industry are now investing in platforms. Of those with a strategy designed to disrupt, 57 percent are builders or owners of a platform business model. Although only 7 percent of Insurance CxOs surveyed currently operate platforms, 26 percent are experimenting with the concept and 21 percent intend to reallocate capital to build or expand platforms.

In every industry, organisations are investing in platforms, and the IBM Institute for Business Value estimates indicate that capital reallocation toward this business model could approach a huge $1.2 trillion in the next two to three years.

As IBM’s Riedel puts it: “Whether or not insurance organizations ultimately choose to operate or participate in new platform business models, they are increasingly likely to be competing with them. As platforms proliferate, every industry seems likely to experience what’s often been called the Amazon effect: the endless evolution and disruption of its markets. The choice of whether to own or participate in a platform, or do both, isn’t something organizations should postpone. Insurers that opt not to participate in a platform will either miss out on the complementary services the platform provides or will have to supply those services themselves, which may prove difficult or impossible.”

The company thinks a development platform for the insurance industry would require these features:

• Automated, agile “build and deploy” processes with the ability to constantly update the software weekly.
• Access to and use of data for personalization and optimization of products and services. –
• Co-creation and joint development with an open and “ready-to-use” environment that has a flexible composition of service packages.
• The ability to incorporate a network of partners so the respective strengths of individual service offerings can be bundled.
• Streamlined integration with minimal setup and scaling challenges through provision of ready-to-use adapters
• Compliance with high security standards, including encryption, authentication and backup procedures

One thing seems clear: digital disruption has belatedly come for the industry and to meet today’s challenges, insurers need to look at ways to increase agility and innovation, so they can better engage with their customers. Traditional insurers need to capitalise on their innate capabilities, such as risk management, and combine them with the flexibility and speed inherent in platforms. To learn how they can do so, see IBM’s whitepaper here.

As a way to diversify risk and optimise workload management by ensuring the right cloud is running the right processes, the approach is a no-brainer: it ensures resilience, flexibility and cost-effectiveness, and most enterprises recognise the benefits.

Yet this cloud approach can come with a lining in need of a distinct polish before it can be called silver: multicloud represents a complex environment to manage; scattered workloads, security gaps and limited visibility on development teams’ output are all operational issues.

This is a problem that is not going away: cloud infrastructure spending has surpassed spending on on-premises legacy IT infrastructure amongst UK-based organisations for the first time, according to a new report by the Cloud Industry Forum this week, and IBM suggests that 80 percent of business workloads have yet to move to the cloud.

Another report the same week by Vanson Bourne found that application mobility across any cloud is a top priority for 97 percent of respondents – with 88 percent of respondents saying it would “solve a lot of my problems.” Additionally, the report found public cloud is not a panacea; IT decision makers ranked matching applications to the right cloud environment as a critical capability, and 35 percent of organisations using public clouds overspent their annual budget. There is huge room for improvement of cloud management here.

The Need for Multicloud Management

Multicloud management is the ability to effectively manage multiple cloud environments (public or private), as if they were a single environment. The need for this is growing fast, given the facts illustrated above.

As Steve Robinson, General Manager, Client Technical Engagement, IBM Cloud puts it: “While IBM has long embraced open source technology, there are some serious IT challenges that Kubernetes alone does not address. Based on our extensive work with enterprises, we have identified three key challenges often found in multicloud enviroments.”

He made the comments as the company released a new Multicloud Manager. This is a cloud-agnostic solution that leverages open-source technology and other existing tools to provide an integrated dashboard. Users can manage their Kubernetes environment and containers where they need to reside — public, private, dedicated or in between.

The tool solves, as Steve Robinson puts it, three challenges.

1. Visibility

With so many clusters across so many environments, many organizations aren’t getting the required visibility into their containerised software. They also can’t see where services are running, how to monitor usage across multiple clouds or how to keep track of clusters as if they were on a single environment. IBM Multicloud Manager, he says, helps improve visibility across all Kubernetes environments. This means that every team has better visibility into the information it requires. Development teams can see deployments, pods and Helm releases. Operations teams can look at clusters and nodes. Security teams can see who has access to which features using a single user interface. All of this means organizations will no longer have to manually check each cluster to see what is happening in their environments. Rather, they can simply view the dashboard within IBM Multicloud Manager.

2. Security and governance

Managing governance with Kubernetes clusters across multiple clouds can be a challenge. Businesses are looking for a way to set consistent security policies across all environments. There is the additional challenge of managing configurations and placing workloads appropriately based on compliance or capability. The tool helps manage environments with a consistent set of configuration and security policies so that an increase in the number of clusters does not result in a change. These policies are enforced at the target clusters. This means they can operate effectively even when connectivity to the management system is lost.

3. Automation

While Kubernetes has some great automation capabilities of its own, some businesses still lack crucial capabilities such as the ability to back up applications, options for managing disaster recovery or the ability to easily move workloads across environments. The IBM Multicloud Manager uses IBM Cloud Automation Manager multicloud automation services to provision, configure and deliver individual Kubernetes clusters as a service in any cloud supported by IBM Cloud Automation Manager.

Ultimately, an agile, multicloud integration architecture should enable users to:

• Manage access to internal and external services with APIs
• Connect on premises and cloud apps to drive business transformation
• Protect APIs, the data they move, and the systems behind them
• Conduct reliable messaging communications across application boundaries
• Move huge amounts of data rapidly, securely and predictably
• Cleanse and prepare data for a consistent view of your business

Those that get this right are going to have a lot more silver lining with their cloud.

The AI Fairness 360 toolkit is an open-source library that provides developers with the means to test AI models for biases, while also providing algorithms to mitigate any issues discovered inside models and datasets.

Alongside the open-sourcing of that toolkit, IBM said it has introduced “new trust and transparency capabilities” on the IBM Cloud that work with models built from a wide variety of machine learning frameworks and AI-build environments such as Watson, Tensorflow, SparkML, AWS SageMaker, and AzureML.

This means organizations can take advantage of these new controls for most of the popular AI frameworks used by enterprises, further identifying AI biases that could skew results in algorithms being used by enterprises.

Release Comes Amid Concerns about Data Bias in “Black Box” Models

Writing in the California Law Review Solon Barocas & Andrew D. Selbst noted of the issue of datasets used in AI training models: “Data is frequently imperfect in ways that allow these algorithms to inherit the prejudices of prior decision makers. In other cases, data may simply reflect the widespread biases that persist in society at large.”

The AI Fairness 360 toolkit tries to address these issue by letting developers access a comprehensive set of metrics for models and datasets that allows them to test their own AI research for inherent biases. Included in the toolkit are the capabilities for the software to inform the user what is biased and why it has been highlighted as issue.

See Also: IBM and Moller-Maersk Aim to Transform Global Supply Chain With Blocks

David Kenny SVP of Cognitive Solutions at IBM said in a release shared Wednesday: “IBM led the industry in establishing trust and transparency principles for the development of new AI technologies. It’s time to translate principles into practice. We are giving new transparency and control to the businesses who use AI and face the most potential risk from any flawed decision making.”

As an illustration of the issues that can result from algorithmic bias, open-source development platform Project Jupyter outlines how a machine learning model can be biased when trying to predict loan repayment outcomes: “Loan repay model may determine that age plays a significant role in the prediction of repayment because the training dataset happened to have better repayment for one age group than for another.”

“This raises two problems: 1) the training dataset may not be representative of the true population of people of all age groups, and 2) even if it is representative, it is illegal to base any decision on a applicant’s age, regardless of whether this is a good prediction based on historical data.”

AI Bias

The IBM Cloud service, meanwhile, which is fully automated, will give enterprises explanations in digestible terms that show what factors were considered in the decision to highlight biases, while also showing what level of confidence it has in the judgement.

All this is displayed in a visualised dashboard on the IBM cloud.

IBM state in the announcement that : “The records of the model’s accuracy, performance and fairness, and the lineage of the AI systems, are easily traced and recalled for customer service, regulatory or compliance reasons – such as GDPR compliance.”

But today’s AI is reaching a tipping point because of improvements in storage technology in terms of both speed and cost, in computing power and in the advanced analytics which can properly exploit both these technologies.

Matt Cox, Nationwide Building Society, said: “There is a new breed of emerging technologies that can take in and make decisions on huge amounts of data, especially unstructured data including images and speech, and there has been a massive step up in the capability to do that in the past few years. Now that we have all this data to apply algorithms to, this is driving a new wave of interest – and all these techniques are possible now at scale, in a way that they just were not before.”

What are the challenges to a successful AI project

The building block for any AI project is data. The old computing adage of ‘garbage in, garbage out’ remains true however big your data lake is. The crucial first step is ensuring the data being used is accessible, clean, useable and reliable. Getting data out of legacy systems and into a functional form for analysis is often the toughest part of any project and is all too easy to overlook.

For financial institutions with systems which have grown organically over time this can require some re-engineering of back-office functions, not simply adding an AI system to the front.

Getting the right skills in place, whether by recruitment or training, is of course vital. But most successful projects also rely on several partner arrangements with other companies. Billions of pounds and dollars have been invested in AI start-ups in recent years, taking advantage of this expenditure will accelerate returns and stop projects “re-inventing the wheel”.

But for financial institutions there is unlikely to a be a single provider of a plug and play solution. A successful project is more likely to require co-operation with several providers.

There also needs to be a recognition from everyone across the organisation that AI is more than just a technology project.

To truly succeed it needs a cultural change across the institution. AI needs to be used by all business units and staff within those units need to be open as to how the technology may impact on their working lives and on the services they deliver.

This process will also need to reassure people that AI can improve their work and is not designed just to replace them.

How is AI being used in financial services right now?

Early stage projects already showing success include customer services and fraud detection. Voice recognition, virtual assistants and chatbots are becoming part of customers’ lives elsewhere and are being adopted by forward thinking financial institutions too.

As people get ever more used to talking to a machine for at least the start of interacting with almost any institution so its use by banks will become more and more accepted.

AI is already playing a big role in detecting cyber-attacks and finding fraud and money laundering attempts. This will only increase because pattern detection is a relatively easy task for AI and extremely useful for spotting dodgy transactions.

What’s next for AI?

Making AI a key part of your financial institution means making sure the staff and customers are equally comfortable with the change. Financial services face a tougher regulatory burden when it comes to data protection and audit trails which will not reduced by intelligent systems. This is especially true because it seems that public attitudes about data privacy and protection are finally toughening up.

That means that AI, or any other form of data processing, must not damage the good reputation most financial institutions have for good data practises. With these provisos in place AI can begin to improve customer services by creating truly personalised offers and automating processing.

Once that is working financial institutions should begin to see faster and better AI-driven decision making and the creation of completely new products and services.

This piece was informed by the research done by a research paper by Finextra titled: “The Next Big Wave: how financial institutions can stay ahead of the AI revolution”.

Who will be the first to implement the new suite of Watson services?

From the newly formed Watson Financial Services division, IBM has released the first suite of services covering regulatory requirements, financial crime insights, and financial risk modelling.

These cognitive tools have been made possible following IBM’s 2016 acquisition of global consulting operation, Promontory Financial Group. Promontory was originally working to provide support to banks dealing with the growing and tightening regulation and risk management within the financial services.

It was the knowledge and expertise accessed in this acquisition that brought life to the new financial services-focussed Watson services, with regulation and risk accounting for two thirds of the suite, and a financial crime tool completing the set. The new tools can be accessed via the IBM Cloud.

Watson’s ability to now provide assistance with regulation and risk may draw the attention of organisationsconcerned by the impending GDPR, and prompt the question of whether the IBM technology could be configured to work in this space.

IBM has also deployed Watson to the Telecoms space after Bluewolf tapped the technology of Einstein and Watson. This involved the combination of IBM Watson APIs with Vlocity and Salesforce Einstein.

Bridget van Kralingen, SVP of IBM Industry Platforms said: “Two generations ago, IBM brought the first computers to the financial services sector, allowing banks and other institutions to foster greater trust in the market by operating more efficiently and accurately”.

“To foster trust today, financial institutions must analyze an industry’s worth of information to monitor risk and compliance. No individual – or team of them – can adequately do this alone, and so once again, IBM is bringing a new type of computing – cognitive computing – to help these professionals operate more effectively.”