Online fraud has gained momentum across many industries, with ecommerce, financial services, social media, and travel tipping the scales from March to October 2020.
In Singapore, for example, ecommerce scams rose by 19.1% to 3,354 cases in 2020. Within the financial services industry, up to 90% of fraud cases in 2020 are attributed to account takeovers (ATOs).
As for social media, there has been a rise of unregulated “finfluencers” who dupe people into making poor personal finance decisions. There is also a group of influencers who use their platforms to blatantly promote guides to frauds, teaching other social media users how to commit these scams on their own.
Finally, even the travel industry is the target of online scams. Despite COVID-19 travel restrictions, airline ticketing frauds have increased. Fraud detectors at ARC identified at least 80 instances of unauthorised ticketing in September 2021 alone, accounting for around $1.2 million.
. . .
The field is ripe for scams – and solutions
With such a vast pasture of opportunities available, fraudsters have sprung up to sow the poisoned seeds of their scams for unsuspecting consumers. And online cheats don’t just stay stagnant and keep employing the same tactics, too – they’re good at adapting to changes, adjusting their strategies to exploit lapses in defences.
Fortunately, all is not lost, as these scams opened a considerable array of opportunities for tech companies to combat them. Dedicated specialists like our portfolio companies IDWall, Authing, and Privy use various techniques like optical character recognition (OCR), Identity as a Service (IDaaS), digital signatures, and machine learning to protect businesses and consumers from fraud.
In this article, we take a look at the solutions devised to fight digital fraud today.
. . .
Optical character recognition (OCR) for quick and safe verification
When onboarding a new client, businesses often need to obtain sensitive personal data. This process can be long and drawn out, leading to errors and even opening loopholes for fraudsters to exploit. Optical character recognition (OCR) reduces the risks.
OCR works on a user-friendly trajectory. Simply scan an image, and – like magic – its text content is automatically read and processed. With OCR, businesses can bypass the manual process of filling out registration forms entirely; all their clients need to do is send over the images of their identity documents for registration purposes.
To further buffer OCR, our portfolio company IDWall in Latin America uses a mixture of documentoscopy and background checks to authenticate the images they receive. Documentoscopy involves criminal forensics, where analysis techniques and specialised equipment are used to test the authenticity of documents.
The OCR process is efficient, too – automation is the key ingredient used to speed up the validation process. As a quick and secure solution, OCR helps businesses onboard more clients without the threat of fraud looming overhead.
. . .
Identity as a Service (IDaaS) to protect enterprise data
The cloud is highly versatile, used in a variety of business applications like storage and processing power. Identity as a Service (IDaaS) is available at your disposal for identity and data protection through the cloud.
IDaaS is very strict in managing the security of your business. It stringently ascertains that your employees match their company profiles before granting them access to sensitive company resources.
IDaaS is applied in three main ways:
- Adaptive multi-factor authentication, where users submit multiple pieces of information about themselves to access a network.
- Single sign-on, a simplified version of #1 that only requires users to sign on once to gain authorisation to all sensitive parts of a network.
- Universal Directory, a centralised cloud-based system that ensures only authorised users can work with restricted data.
IDaaS is offered by our portfolio company Authing. Based in China, Authing strongly believes in adopting an identity-centric approach to security. With this approach, your business will be better poised to safeguard its sensitive systems and confidential information from the prying hands of fraudsters.
. . .
Digital signatures for secure signings and verifications
Gone are the days when you would hand-sign documents for approvals and verifications. Digital signatures are the new way to sign documents, and these are powered by mathematical algorithms that periodically validate their authenticity and integrity.
Only verified signatures are permitted to authorise sensitive documents with these algorithms at work, preventing fraudulent signings of malicious documents from occurring.
Our Indonesia-based portfolio company Privy offers digital signatures as a solution. Known as PrivySign, this solution utilises a blend of asymmetric cryptography and public key infrastructure (PKI) to create unique signatures tied to verified identities.
. . .
Machine learning to stamp out frauds
Machine learning isn’t a new-fangled concept; it’s a branch of artificial intelligence (AI) that works like its name suggests: machines are trained to learn. Algorithms use automation to recognise patterns and learn how to perform tasks without being explicitly programmed.
As time goes on and the algorithms continue to process information from previous computations, they can independently produce consistent, reliable, and repeatable decisions and results with little to no human intervention.
Commonly used in sectors like retail and financial services, machine learning is also a viable tool against scammers. Fang Yu, co-founder of fraud detection company DataVisor, uses machine learning called “unsupervised detection”.
Unlike other fraud-fighting solutions that tend to react after frauds occur, unsupervised detection anticipates them ahead of time. This method also predicts the tendency for fraud patterns to evolve at the drop of a hat, capable of detecting potential security breaches in real-time. As Fang asserts,
“Fraud patterns sometimes evolve overnight or over a few hours. If you’re always writing rules or training machine learning solutions, it’s actually very slow and always reactive.”
. . .
The fight against fraud is not over
While the solutions presented so far have proven to be formidable foes to digital frauds, fraudsters are learning how to work around existing technology and evolving their methods.
Fraud protection companies will, of course, work to respond and adapt to these challenges, but your first line of defence is yourself. Here are some quick tips to protect yourself from falling victim to digital fraud:
- Use strong, unique passwords for different online accounts.
- Do not freely share your credentials with every site you visit.
- Do not click on links from unverifiable sources.
- Only make online purchases from reputable websites.
- Constantly review your credit score.
As the adage goes, it’s better to be safe than sorry. Arm yourself well, and you will significantly reduce your chances of running headlong into a fraud.