Navigating Secure Authentication Across the Wild West of the Internet

By Ralf Gladis, CEO, Computop

Remaining secure while benefiting from all the Internet has to offer is what we, as consumers, seek.  Which is why secure identification and authentication are increasingly vital. The transfer of sensitive data, including payment transactions, must be performed with as little risk as possible to protect not just customers, but service providers too, against the theft and misuse of valuable information.

Currently the largely uncharted digital world can feel like the wild west when it comes to security. The combination of user name and password is comparatively insecure and alternative, more sophisticated methods are competing for the best positions on the market. We have yet to see the solid establishment of a standardised and generally accepted system for secure identification and authentication, but in the meantime, it’s important to understand the landscape.

Identification and authentication

These terms are often used synonymously, but they describe two different processes. Identification is when a person proves their identity to an authority or entity to which they were previously unknown. This occurs, for example, via conventional registration with an email address and password, which is sufficient for many services. For more sensitive applications such as payment transactions or banking, on the other hand, there are more sophisticated identification processes such as Postident and WebID. These use significantly more complex methods to check whether a person corresponds to the identity he or she claims to have.

Authentication, conversely, involves recognition. After a user has identified themselves and registered, they must log in and for this and all subsequent uses, be authenticated. The usual pairing of user name and password entered during registration are typically used for this purpose.

However, this method has long been criticised because compared to other processes, it is relatively insecure – particularly when the user’s email address also serves as the username.In addition, many customers consider password management to be tedious which means that instead of using complex letter and number combinations they resort to an easily memorable code based on birth dates or family names. Unsurprisingly this is easy to crack and presents a high security risk.

Social-login solutions from sites including Facebook, Twitter, LinkedIn or Google apply once the user has registered, which means that website operators don’t store the log-in data on their servers. Instead, authentication is performed via the plug-in of the relevant site, making it harder for criminals to steal login information and hack the account. Many platforms offer optional two-factor authentication via mobile phone number, which makes them significantly more secure than with just one password or with access via a password manager but this is not compelling to users who are often reluctant to give the sites their mobile numbers. To address this, Apple has just launched a new privacy tool, ‘Sign In with Apple’, which uses customer’s Apple ID rather than their email address, to verify credentials.Meanwhile, the social media giants still lack the final proof of identity which can only be provided by service providers who identify their users in a lawful manner and then securely link the information to an authentication process. They are then able to link each authentication to the identity behind it.

Processes such as ‘Identity management’ and ‘Postident’ in Germany or BankID in Sweden are state-regulated identity verifications which meet high security requirements. Regulators recognise them as equivalent to identification through personal presence. These processes must fulfil specific regulations in the relevant countries to be accepted as secure by the government. For example, in Germany, the identification must be carried out by service provider employees who are in access-protected rooms. Additional requirements include uninterrupted video identifications in real-time and adequate image and sound quality.

PKI

Identification and authentication can be encoded asymmetrically via a private and a public key. This relies on a certification authority (CA) which verifies public keys and issues digital certificates for them. The key pair is usually generated on the device or smartcard of the user. The private key always stays with the user, while the public counterpart, which has been signed by the CA, is submitted to the service for which they are registering. For authentication, the service provider sends the user a calculation which they can only solve if they possess the private key -this is the central security element. Only the service with the matching public counterpart which has sent the request is able to check the solution. The security provider only issues it in a protected environment. This may be, for example, a protected hardware sector in an iPhone. They sign the public key with a certificate authority (CA). As with the SSL certificate of a website, the certificate is verifiable for any outsider and is generally issued for the email address of the user. The authenticity, confidentiality and integrity of messages are guaranteed. If the issuer of the certificate has checked and verified the identity of the user, the user can use it to sign documents in accordance with signature legislation. The digital signature thus replaces the ‘wet’ signature.

With clearly and lawfully identified customers, data centres and smartcards, banks have the ideal conditions to work PKI, but they are not widely adopting it at the moment.

FIDO

To reduce password reliance, the FIDO Alliance (Fast IDentity Online) is establishing public and licence-free industry standards for global online authentication. Like PKI, FIDO uses a pair comprising a public and a private key. However, the duo is generated through the FIDO authenticator, a protected software area in the smartphone, and this supports user verification which takes place every time the key is used, for example via biometric methods such as iris or fingerprint scans.

Almost all operating systems offer suitable interfaces.For web services, FIDO is a convenient and secure option for authentication which they can integrate easily using a FIDO server. With web authentication, the W3C consortium (a body for the standardisation of technology on the Internet) has adopted an authentication standard for web browsers with a FIDO connectionalready being used in Microsoft Edge, Google Chrome and Mozilla Firefox and is being considered by Apple for Safari. It enables password-free authentication in the browser. Users can log into websites using their fingerprint or via face recognition, rather than a user name and password. Manufacturers can use the API for smartphones and tablets with a fingerprint sensor (e.g. Touch ID) or face recognition (Windows Hello, Face ID). The biometric user data therefore remains in a secure area of the device, where it never leaves.

FIDO & PSD2

The revised Payment Services Directive (PSD2) aims to make electronic payments in Europe more convenient and secure by requiring stronger authentication. To this end, it must manage at least two of the following three factors:

  • Knowledge: Information that only the user knows (e.g. password).
  • Possession: Something that only the user owns (e.g. smartphone).
  • Inherence: Something that is a personal or physical aspect of the user (e.g. fingerprint)

FIDO enables secure authentication without passwords. A separate key pair exists in the FIDO authenticator for every service to which a user logs in. The authenticator represents the factor of possession. A smartphone combines it with the factors of knowledge or inherence. The FIDO authenticator builds on this as it can only be activated via a PIN, fingerprint sensor or facial recognition. Computop’s FIDO solution will enable merchants to provide their customers with a biometric login to their webshop account but also to authenticate their payment in a safe and quick process that will be beneficial for the shopping experience. It is important to note that biometric data is stored in a highly encrypted secure element on the customer’s device and is never handed over to the service provider.

Blockchain

Blockchain is a continuously expandable list of records (blocks) that are chained through cryptographic techniques, whereby each consists of a cryptographically secure hash of the previous block, a timestamp and the transaction data. If someone wishes to manipulate a block, they must also change all the following blocks, a costly exercise making this type of fraud unprofitable and the technology tamper-proof.

Identity providers and brokers can store the customer’s identity information along with the offered trust level, the cost of an identification and the delegation address for the authentication in the blockchain. The end customer clicks on the central checkout button on the e-tailer’s website and enters their email address there. The system checks if a local authentication is available for the customer – either via a FIDO biometric solution or a user name and password combination. It then asks the blockchain if there is a registered identity provider for the customer. If not, they can register locally. The identity provider delivers the identity information and confirms the successful authentication.

Summary

PSPs like Computop have to be prepared for the challenges that the future presents. As we’ve seen, numerous initiatives will ensure secure digital identification and authentication in the future. Standards such as FIDO set the course for future business and government communication which gets by without passwords. Google, HUAWEI, Intel, Lenovo, Microsoft, Samsung and others are working on the aspects of functionality and convenience. Retailers can decide if they identify and authenticate customers using their own system or instead work with service companies and full-service identity providers. However, development on the market shows one thing clearly: the reign of passwords is coming to an end.