Decentralization in the COTI payments network, KYC/AML and Trust Score cases
Written by Anton Suslonov, data scientist, specialist in deep learning, neural networks and natural language processing
This Monday, April 2, 2018, we will be hosting a live AMA with Anton about decentralization in the COTI payments network, KYC/AML and Trust Score use cases. Anton will be discussing the influence of these issues on the future of online payments, so be sure to tune in. The AMA will take place in COTI’s Telegram group from 4:00–5:00 pm (+2 GMT).
Imagine the world 20 years from now. How do you foresee people paying for goods and services? The world will likely not be using coins, cash, cards and cheques anymore, but digital currencies, which are faster and easier to use, imparting economic freedom to societies around the world. In this article, we’ll be exploring the various types of networks and their inherent features.
Any type of electronic money, or other means of payment, is stored in a specific type of a network, each with its set of idiosyncrasies that affect efficiency, reliability and safety. We will be considering three network types (and countless variants and mixed models), such as centralized networks, permissioned networks and decentralized networks.
In a centralized network, all network nodes are directly or indirectly connected to a central server, while network resources are controlled from a single location.
Centralized networks typically offer users high performance, high reliability and an extensive set of useful features depending on network architecture. These networks, however, are not without fault, as growing numbers of people prefer not to provide central authorities with undue control over their data.
Such networks also face cyber attack vulnerabilities, whether by third parties or network administrators and authoritative bodies. Compounding this issue is the fact that centralized networks are not typically based on open source software, which deprives users from viewing the code.
The best example of a centralized payment system is VISA. It consists of processing centers, member banks, non-member bank issuers, POS terminals, ATMs and more. This type of payment system is efficient, but very expensive and rarely open to innovation. Smart cards, for example, took decades to be deployed.
Distributed permissioned networks
Ripple and IBM Hyperledger are both examples of distributed permissioned networks. Such network nodes operate independently and run the same open source code, while critical functions are controlled by a central authority. Centralized functions typically include node synchronization, promotion and demotion.
Distributed permissioned networks usually exhibit outstanding performance and stability, particularly for international letters of credit and interbank payments.
Decentralized networks were designed to overcome known central authority vulnerabilities. The Internet, for example, was designed as a decentralized network with certain permissioned aspects, such as IP address designations and DNS servers.
In FinTech applications, decentralized network nodes should be ideally suited to process transactions in an equitable and trustworthy manner, which can be achieved using a consensus mechanism.
Decentralized networks are exempt from the autocratic influence of central authorities. The primary challenge is to prevent whole network failure that could be brought on by cyber attacks. As such, nodes are tasked with performing proof of work (PoW) in order to reach consensus.
COTI is a decentralized payment network that aims to be the most scalable and efficient system around. COTI’s Trustchain Consensus Algorithm enables Full Nodes to independently aggregate their portions of the DAG. Nodes verify one another without the need for block synchronization. The Double Spending Prevention Nodes work in parallel, imparting reliability and security to the system without deadlocks or bottlenecks.
Trust Score values enable the system to run efficiently while equitably distributing new transactions within the DAG. For this reason, COTI’s Trust Score Algorithm is fundamental to the functioning of the COTI payments network — it is unbiased, decentralized and open.
KYC/AML requirements in decentralized networks
Know Your Customer (KYC) and anti-money laundering laws (AML) are a strict and sophisticated set of rules introduced by U.S. lawmakers (see USA Patriot Act) and the Financial Action Task Force (FATF). COTI is taking a proactive approach to digital currency-specific regulations by adhering to KYC/AML procedures.
The aforementioned regulations, however, limit the use of cash in transaction settlements. According to the SEC and CFTC, “market participants should treat payments and other transactions made in cryptocurrency as if cash were being handed from one party to the other.”
The primary focus of KYC/AML procedures is identity verification and proof of residence. This necessitates the need to process and collect real world documents and offline confirmations.
The COTI network’s KYC/AML servers are squarely tasked with maintaining KYC/AML procedures, although they do not store information concerning transactions and wallets.
Trust Scores in the COTI network
Documents containing personal data must be securely stored. The challenge, however, is maintaining uniformity and deferring system manipulation in an effort to sustain the efficiency of the Trust Chain consensus.
Our decentralized solution to this challenge:
In the COTI payments network, the Trust Score Node receives data from the KYC/AML server. This node does not have access to a user’s personal data, which enables Trust Score Nodes to be operated by network participants.
Users are incentivized for equitably running Trust Score Nodes by earning fees for their efforts. By comparing Trust Scores assigned by different nodes to the same user, the network will be able to effectively detect any fraudulent attempts.
The live AMA with Anton will be taking place 2 April, 2018 from 4pm — 5pm (GMT +2) in the COTI Telegram group. See you there!