Most companies big and small tackle identity fraud daily and have come to rely on a fleet of tools, including multifactor authentication and CAPTCHA (completely automated public Turing test to tell computers and humans apart) codes, to help identify potential identity fraud. While these tools help to some extent, they don’t catch everything. According […]

Mercator Advisory Group releases a new research report that examines the impact of hidden biases in ML and Artificial Intelligence--and how to avoid them. BOSTON, Oct. 12, 2020 /PRNewswire-PRWeb/ -- AI models reflect existing biases if these biases are not explicitly eliminated by the data scientists developing the systems. Constant monitoring of the entire operation is required to detect these shifts. The remedy for such lack of focus is training. Mercator Advisory Group's latest research Report, Tracking Mistakes in AI: Use Vigilance to Avoid Errors, discusses modes in which data models can deliver biased results, and the ways and means by which financial institutions (FIs) can correct for these biases.

AI has been used to create deep fake images, voices and videos. Researchers believe that it may soon be impossible to tell the difference between a real person and a fake. "Criminals used artificial intelligence-based software to impersonate a chief executive's voice and demand a fraudulent transfer of €220,000 ($243,000) in March in what cybercrime experts described as an unusual case of artificial intelligence being used in hacking. The CEO of a U.K.-based energy firm thought he was speaking on the phone with his boss, the chief executive of the firm's German parent company, who asked him to send the funds to a Hungarian supplier. The caller said the request was urgent, directing the executive to pay within an hour, according to the company's insurance firm, Euler Hermes Group SA. Euler Hermes declined to name the victim companies. Law enforcement authorities and AI experts have predicted that criminals would use AI to automate cyberattacks. Whoever was behind this incident appears to have used AI-based software to successfully mimic the German executive's voice by phone. The U.K. CEO recognized his boss' slight German accent and the melody of his voice on the phone, said Rüdiger Kirsch, a fraud expert at Euler Hermes, a subsidiary of Munich-based financial services company Allianz SE. Several officials said the voice-spoofing attack in Europe is the first cybercrime they have heard of in which criminals clearly drew on AI. Euler Hermes, which covered the entire amount of the victim company's claim, hasn't dealt with other claims seeking to recover losses from crimes involving AI, according to Mr. Kirsch."

There are times when the terms used to describe technologies just don't tell the entire story. Machine learning and artificial intelligence (AI) are two recent examples. A topic of countless news stories and plenty of futuristic speculation, machine learning and AI are often used as the main characters in scary dystopian tales of robots that gain self-awareness and then use their powerful computer brains to take over the world. That's fine as fodder for science fiction tales, but it also can serve as a major (not to mention unnecessary and untrue) obstacle to harnessing the legitimate power of AI and machine learning to solve real-life business challenges today. Put another way, machine learning and AI have an image problem – one that the technologies can't solve for themselves because, well, they're not the omnipotent robots of our imaginations.

Luckily I've had the opportunity before for you and I to have a couple of conversations about DataSeers, and I was wondering if you could give our audience a brief overview of your organization and its role within the payments industry? As the name suggests, we are data seers, which means we see through data. If you look at the payments industry today, it is generating large volumes of data. It's creating a large variety of data because payments are very different when they come from different providers, different processors, and so on and so forth. And it's also coming very fast, so that volume, velocity, and variety creates a toxic mix for banks and other companies in the payments ecosystem to handle.

Artificial intelligence is one of the most compelling areas of computer science research. AI technologies have gone through periods of innovation and growth but never has AI research and development seemed as promising as it does now. This is due in part to amazing developments in machine learning, deep learning, and neural networks. Machine learning, a cutting-edge branch of artificial intelligence, is propelling the AI field further than ever before. While AI assistants like Siri, Cortana, and Bixby are useful, if not amusing, applications of AI, they lack the ability to learn, self-correct, and self-improve.

As many who follow the financial services industry will recall, during and after the financial crisis of 2008-2009, a number of events and structural changes occurred. One of those changes involved former investment banks being either acquired (i.e.; Merrill Lynch, Bear Stearns) or converting to a financial holding companies (FHC), which allowed for easier capital access but also subjected them to closer regulatory scrutiny. One of the FHC conversions (through the Gramm-Leach-Bliley Act of 1999, essentially the reversal of 1933 Glass-Steagall Act) was Goldman Sachs. This piece, appearing in Tech Republic, uses the Goldman Sachs digital banking platform Marcus as a proxy to make a point about the changing consumer banking landscape.

Conversational AI for banking – by its very nature – is a constantly-evolving technology. Based on simple human conversations, this amazing science continually refines and optimizes itself each time it interacts with a user. We call this "training the bot", an ongoing process that relies on plenty of data and user interaction to deliver consistently accurate responses from the Finn AI virtual assistant. With such cutting-edge technology, the question I get asked most often is: what's next? At Finn AI, we've recently partnered with Visa to explore what's next.

"In hiring, the bank wants to use AI to help source the right candidates. "There's a chance AI models will be biased," said Caroline Arnold, BofA's head of enterprise technology (which includes HR tech). "You might say, who's going to be successful at this company? An AI engine could find that people who golf are going to be successful at the company. On the other hand, using those same techniques can remove bias if you have the model ignore some of these things that are indicators of different groups but go on to the meat of the profile of the person and understand it in a deeper way." Arnold believes an AI engine can never be the final say in who gets hired. Mehul Patel, CEO of Hired, a technology company whose software uses AI to match people to jobs, agreed that AI and humans have biases. "The good news about AI is, you can fix the bias," he said. "We will boost underrepresented groups.

IBM and Apple have agreed to link Watson's machine learning development platform so that trained models in Watson can be executed on the iOS platform, but apparently only within the Mobile First initiative which restricts availability: "Integrating Watson tech into iOS is a fairly straightforward workflow. Clients first build a machine learning model with Watson, which taps into an offsite data repository. The model is converted into Core ML, implemented in a custom app, then distributed through IBM's MobileFirst platform. Introduced at the Worldwide Developers Conference last year, Core ML is a platform tool that facilitates integration of trained neural network models built with third party tools into an iOS app. The framework is part of Apple's push into machine learning, which began in earnest with iOS 11 and the A11 Bionic chip. "Apple developers need a way to quickly and easily build these apps and leverage the cloud where it's delivered," said Mahmoud Naghshineh, IBM's general manager, Apple partnership. On that note, IBM is also introducing IBM Cloud Developer Console for Apple, a cloud-based service that simplify the process of building Watson models into an app. The arrangement allows for back-and-forth data sharing between an app and its backbone database, meaning the underlying machine learning model can improve itself over time if the client so chooses. Users can also tap into IBM cloud services covering authentication, data, analytics and more. As you run the application, it's real time and you don't need to be connected to Watson, but as you classify different parts [on the device], that data gets collected and when you're connected to Watson on a lower [bandwidth] interaction basis, you can feed it back to train your machine learning model and make it even better," Naghshineh told TechCrunch.