Insurance Fight insurance fraud: data sharing with blockchain technology

Transcription

1 Insurance Fight insurance fraud: data sharing with blockchain technology Establishing trust, accountability and transparency in data sharing

2 2 Fight insurance fraud: data sharing with blockchain technology Executive summary This paper introduces the concept of blockchain and its application in the handling of insurance claims and managing fraud in various lines of business in the insurance industry. The problem: The bad guys know more than the good guys Each insurance company has a narrow lens. Criminals exploit gaps in visibility. Here are a few examples: The same device is used to create multiple policies. Synthetic IDs are used to create multiple policies, identifiable only by looking at wider data set. The same parties involved in multiple fraudulent claims have different roles (for example, customer, suspect, victim, witness, provider) The same patterns are used in multiple claims across insurers by the same or different people (for example, text phrases or rings) Each insurance company invests heavily in similar public and subscription data to inform fraud prevention and investigation. In some jurisdictions, they pool resources to lower cost and improve insight. Derivative intelligence using third party or external data Cost considerations by outsourcing and fully run special investigation units Conclusion: There is value in sharing intelligence to expand the observation space. So why hasn t this happened? Insurers have tried to share intelligence in past but face technical, legal, and commercial challenges. Here are a few examples: Sharing personally identifiable information data Sharing data across national borders Unwilling to share sensitive loss information with competitors or third parties Who manages the shared data? Who can be trusted? Larger institutions contribute more and get less than smaller institutions (the free rider problem) Hypothesis: Could blockchain technology help make fraud intelligence sharing viable? The obvious advantages in creating sharing platform on blockchain will include immutability, accountability and transparent compliance. Subrogation could be a good example where this sort of feature can become very useful. If a network of insurers can collaborate to develop a blockchain network for the ecosystem, as mentioned in the problem statement above, a large number of distributed processes can be housed on the network, with no single owner and full transparency. Introducing blockchain technology Blockchain is an append-only ledger enabling digitally signed transactions and embedded business logic that is replicated and shared across legally separate organizations. The promise of blockchain is to remove the need to trust a third party by trusting the network-agreed data set contained in each block and visible to all parties in the chain. IBM has set a goal to accelerate making blockchain technology ready to support enterprise use cases by contributing to open source projects that aim to make blockchain real for business. Note: IBM is not modifying bitcoin or creating its own blockchain fabric. Rather, IBM is working to donate some of its best work on the subject of replicated ledgers, cryptography, security, distributed data and other key blockchain-related competencies to the open source community. It is hoped that such donations will speed the maturity of blockchain for business. How is blockchain related to Bitcoin? Blockchain makes the popular cryptocurrency, Bitcoin, possible, ensuring that coins passed between parties are cash-like: unique, immutable and final.

3 Insurance 3 The way Bitcoin implements blockchain makes bitcoins an unregulated, censorship-resistant, shadow currency. IBM has no interest, and does not support, unregulated currencies that operate outside regulated financial markets. The blockchain is seen as a main technological innovation of Bitcoin, since it stands as proof of all the transactions on the network. A block is the current part of a blockchain, which records some or all the recent transactions, and once completed goes into the blockchain as permanent database. Each time a block gets completed, a new block is generated. There is a countless number of such blocks in the blockchain. To use conventional insurance as an analogy, blockchain is a like a full history of insurance transactions. Bitcoin transactions are entered chronologically in a blockchain just the way insurance transactions are. A block, meanwhile, is like an individual. What does blockchain technology offer insurers? Blockchain-based systems could help radically improve the insurance industry. But its impact could be much broader. Insurance claims processing and settlement, are areas where a customer buys a policy and the policy sets out rules under which, the customer gets a coverage. The level of expertise required for claim adjusters or the investigators, who investigate, negotiate and settle claims, varies directly with the nature of loss, under the jurisdiction whose law applies to the contract creation, interpretation and enforcement, one of principal consumer protection issues within the insurance pricing, underwriting, and claim settlement process, the rules, aka the rules-based claims processing or otherwise a voting based claims processing, mostly followed by mutual insurers (peer-to-peer). In the claim process, a person insured and other parties need to send several documents (data) some of which might not be available in digital format. The settlements are also subject to interpretation of the rules that apply to the product and policyholder situation. This leads the claim processor to apply rules and make decisions before paying out the claim. This goes into building incentive models to enable groups to decide which claims should be paid out and which claims should not be paid out. Further, in lines of insurance where losses require onsite examination, claims adjusters or an investigator may be sent to conduct an investigation. Growing numbers of companies now specialize in offering claims investigation services for insurance companies. Such decentralization of the insurance production cycle is becoming increasingly common, creating a market place, through the lowering of the risk and cost, enabling a breaking of existing monolithic structures, where a decentralized model has become predominant. First, this creates the obvious question of mutual trust between parties. The insurers and their network must be able to collaborate within the ecosystem in a transparent way to prevent fraud. Blockchain may be able to prevent fraud because the transactions are transparent and there can be fewer forgeries or manipulation of policies and contract. Second, if these features could be built into the existing market infrastructure, it would allow for the auto-execution of coded logic embedded into smart claims. All transactions written to a blockchain are visible to insurers, brokers, third-party administrators, service providers shared cross multiple locations and impossible to alter or delete. As the basis for establishing consensus and enabling a contract between multiple insurers with an intermediary will be very useful as it creates a permanent audit trail. Mutual consensus verification protocols allow a network to agree on updates to the database collectively, with a certainty that the overall data set remains correct at all times without the need for a central governing authority. Blockchain transactions can already be reverse engineered to allow a pretty good guess at the transactor. Another blockchain virtue is the decentralization of customer databases and a federated identity infrastructure that allows an individual to identify information online without having to supply personal information (for example, a driver s license number to multiple insurers).

4 4 Fight insurance fraud: data sharing with blockchain technology Rather than being held in any number of central databases, sensitive data is encrypted and distributed and a blockchain is used as an audit ledger to re-authenticate the user. Can blockchain reduce fraud directly? Blockchain may be useful in reducing fraud related to the integrity of a policy or claim or vehicle (any asset). By maintaining the integrity of the asset through various owners, blockchain technology can minimize counterfeiting, double booking, document or contract alterations. It has the potential to be used for identity management, though this concept has a number of issues that need to be worked out. However, use of the blockchain does not mitigate the risk associated with the majority of first-party and third-party frauds. Fraud detection and investigation systems will still be required. This is true whether the blockchain is managed by an individual financial institution, a group of financial institutions operating a network or utility, a third-party processor, or even IBM itself. (You would need to run something like Counter Fraud Management System.): Identity and relationship resolution: Must know for each party who is who and who knows who Behavioral analytics: Must monitor behavior for parties, policies, and devices across channels and lines of business Descriptive and Predictive Analytics: Must be able to identify clusters, anomalies, train models Unstructured Data Analysis Link Analysis Alert and Case Management Regulatory Reporting Loss Reporting Use blockchain technology to share fraud intelligence among institutions Storing and reconciling data sets of financial obligations and ownership forms are the core of insurance claims operations. The current methods are highly complex, use fragmented IT and data architectures and suffer from a lack of common standards. This creates a continual need to reconcile data with massive systems and process duplication, leading to high costs and protracted time to execute tasks. Blockchain can be used to manage shared intelligence and could be used to drive efficiency in insurance claims processing independent of fraud, thus bringing a structural change to the industry. If the insurers can record transactions on a blockchain at each point in the transactional lifecycle, from seeking a quotation to binding a policy contract, the immutable life record of that policy or the policy holder can be traced. This does bring a degree of simplicity to the underwriting process and has the potential to reduce fraud. Insurance claims, is a distributed process that involves many parties, insured, insurer and third parties, often with insufficient transparency to the participants. By using blockchain, insurers can create receipts at different points in the claims process, resulting in an immutable, auditable record of all claims activities, which can be revisited by all parties including the regulators. This could lead to lower transactional cost, lower transaction risk and trustless computation. These new technologies make it possible for a group of independent parties to work with universal data sources, automatically reconciling between all participants customer, broker, insurer, co-insurer and reinsurer. All could have a distributed, single view of the entire exposure data chain. In principle, any stored data record could be represented on a blockchain, from ownership of assets to contractual obligations, credit exposures or static data.

5 Insurance 5 1. Claim transaction 3. Counter fraud servicing Faceted ledger stores claims details and transaction history (in effect assumes the traditional custodian role) Insurance fraud bureau (IFB) is virtual layer coordinating the role of custodians Reference database (third party) Scoring services Pre-claim verification and affirmation of case Faceted search ledger e.g. Derivative assets Investigation services Insurer A Central reporting ledger Fraud management and KPI Cases Verification Transaction hash Counter fraud platform Insurer B Cases (e.g. investigation folio) Securely records ownership rights and encodes claim transaction details Fraud network analysis ledger e.g. Identify verification e.g. Entity verification Risk management Interoperable network analysis ledger enables near real-time monitoring of customer behavior Finance and accounting 2. Case transaction Insurer A Case ledger Investigation ledger Case management Cases Smart cases Counter fraud platform Cases programmed to execute automatically e.g. Cases as assets Interoperable investigation ledger enables near real time, efficient detection of cases for scoring Insurer B Case ledger stores smart cases and execution algorithms Reference data sources are agreed rules for scoring detection and triage positions and obligations (can be distributed ledger or regular data sources) Figure 1. Integrating counter fraud platform with blockchain

6 6 Fight insurance fraud: data sharing with blockchain technology A multitude of data types can be hashed, encrypted and entered into a ledger to create richer data sets than are available today. For example, ownership data could be entered showing multiple levels of beneficial ownership, collapsing the hierarchies that exist in various custody arrangements. Participants store distributed records locally as their golden source of information. Many of their existing systems that are currently used to track and maintain their records of holdings and transactions could be retired. The need to interrogate centralized databases or send messages to other participants to ensure data alignment is removed. Criminals could no longer crash for cash, or exploit the current challenges of sharing data unless their methods for obscuring identities became significantly more sophisticated. Many might be skeptical that anyone will let a decentralized authority manage identities. A common claims-handling platform would still make it possible for individual insurers to compete for customers, offering a range of products and prices by virtue of the smart contracts they set up. Some of these benefits might be achievable with existing technologies, or indeed with no actual technology at all. Adoption of blockchain technology will rely on aligning industry standards for the process, data terms, contractual documentation and more. Regardless of technological innovation, this standardization can improve settlement times and help control costs even using existing market practices and infrastructure. A central authority could maintain a single, universal source of the truth database recording asset transactions which all participants use as their golden source. Counter fraud platform If...and then... Smart claims Transaction block This is essentially an expansion of the role taken by the Insurance Fraud Bureau in a traditional infrastructure, a central hub for sharing insurance fraud data and intelligence, using their unique position at the heart of the industry and unrivalled access to data to detect and disrupt organized fraud networks and therefore the claims management process. Adopting a common blockchain could create a step-change in claims handling by being more efficient and streamlined, resulting in an improved customer experience. Such an approach could also help to further reduce, if not entirely prevent, fraud if identity management was also enforced on the blockchain. Shared ledger Figure 2. System diagram of sharing intelligence using blockchain to counter fraud Moreover, a blockchain could allow the industry as a whole to streamline its processing and offer a better user experience for customers who have to make a claim. Simultaneously, storing claims and customer information on a blockchain could help minimize fraudulent activity. It would certainly make it more difficult for criminals to attempt to claim more than once, if not mask their identities. Indeed, in many respects, with projects like

7 Insurance 7 the IFB now long-established, the general insurance industry faces a smaller cultural and organizational challenge to overcome than does the banking industry and other sectors. Where can you start? Any journey towards claims and fraud management systems based on blockchain technology will be a step-by-step adoption rather than a big bang reorganization. The system is simply too big, complex and important. Individual use cases for the technology need to be identified and solutions developed. Initially, these use cases need to be standalone. They can be adopted within or alongside today s architecture without being dependent on a critical mass of assets already being on blockchains. Working on a concrete proof-of-concept or uses cases is the starting point. Use Case 1: New blockchain to share data between participants to counter fraud: The process of a business verifies the identity of its customers. Insurers to ensure their proposed agents perform Know Your Customer (KYC) checks, brokers or intermediaries are compliant against anti-money laundering, anti-bribery, terrorist financing including the customer s integer, prevent ID thefts and fraud. Use Case 2: New blockchain to store and trace identities for fraud prevention: Provide an inter-insurer blockchain to support cross-border, network linkage analysis, also called social network analysis, has proven particularly beneficial against insurance fraud. Identity ledgers that store and trace identities of people or businesses can help reduce cost of verification and reduce fraud and identity breach costs over time. Use Case 3: New blockchain to store risk intelligence to reduce fraud: Risk ledgers that store and trace information about the risk being insured (for example, property titles, land titles, precious gem fingerprints and high-value asset fingerprints) can help reduce cost of verification and reduce fraud costs over time. Other possible relevant elements that can be housed on a blockchain are digital rights management and Proof-of-Authenticity to counter fraud. How can IBM help? IBM announced a new framework for securely operating blockchain networks. 1. A creation of a catalog of processes that can use the blockchain features. 2. Define a security framework for blockchain on the cloud. Built to help businesses quickly host secure, tamper-resistant networks and scale to thousands of users, IBM Cloud will allow production blockchain networks to be deployed in minutes, running signed, certified and tested Docker images with dashboards and analytics, as well as support. These new cloud services have been optimized for cloud-based blockchain networks by providing an auditable operating environment with comprehensive log data that supports forensics and compliance. Tamper-resistant storage of cryptographic keys and complete protection around the cryptographic module detects and responds to unauthorized attempts at physical access. IBM Cloud services also enable blockchain peers to run in protected environments, to prevent leaks through shared memory or hardware. 3. IBM Blockchain for Hyperledger is now available on DockerHub. For organizations that need the flexibility to run blockchain on different cloud servers or devices, a signed, certified distribution of the IBM code submission to the Hyperledger project is now available on DockerHub. Ongoing updates will provide new features including dashboards, analytics, chat support and exclusive network services. 4. IBM Blockchain for Bluemix provides developers new flexibility. For application developers who want to get a blockchain environment running almost instantly and start building applications, the beta release of IBM Blockchain on Bluemix provides access to the very latest Linux Hyperledger code, updated as the code continues to emerge. Whether on DockerHub or IBM Cloud, IBM Blockchain supports multiple industries and diverse use cases.

8 For more information To learn more about the IBM framework for securely operating blockchain networks, please contact your IBM representative or IBM Business Partner, or visit the following websites: ibm.com/insurance and ibm.com/blockchain About the authors Dr. Indranil Nath is IBM Vice President and Executive Partner, Insurance, Industry Solution & Sales, Europe, and has performed a number of major engagements particularly relevant to digital insurance, information technology delivery, process transformation and automation and operational excellence in the insurance industry. He also leads the Insurance Counter Fraud initiates in IBM Europe. He is a Fellow of the Chartered Management Institute, a Chartered Fellow of BCS, the Chartered Institute of IT, and a member of the Chartered Insurance Institute in the UK. Indranil has many articles to his credit in the areas of training and development, business and technology management and software engineering in the insurance industry. Indranil can be reached at indranil.nath@uk.ibm.com Other contributors are, Mr. Frank Pinder, Global Head of Insurance Counter Fraud, Global Business Services and Mr. Austin Wells from the Offering Management Counter Fraud and Financial Crimes, IBM Analytics Copyright IBM Corporation 2016 Global Business Services Route 100 Somers, NY Produced in the United States of America September 2016 IBM, the IBM logo, ibm.com, and Bluemix are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the web at Copyright and trademark information at ibm.com/legal/copytrade.shtml This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates. It is the user s responsibility to evaluate and verify the operation of any other products or programs with IBM products and programs. THE INFORMATION IN THIS DOCUMENT IS PROVIDED AS IS WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided. Statements regarding IBM s future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only. Please Recycle IUW03049-USEN-00