How AI fights against financial criminals and money launderers?

As criminal methodologies are growing more advanced, the fight against money laundering is becoming a huge challenge for all the financial institutions around the world. Therefore, it becomes necessary to put in AML (Anti-Money Laundering) measures. As AML requires to deal with a huge amount of customer data, they are turning to AI and Machine Learning, to help them identify and detect money laundering activities.

AI performs AML tasks faster than a human employee and also, through machine learning it possesses the capability to modify new threats and detect new money laundering methods. It ensures that financial institutions are able to adjust quickly to different regulatory environments.

When transaction data of a customer is incorporated into an AML program, AI and machine learning models analyze the behavior to make predictions and perceptions about that customer in the future.

How are AI and Machine Learning advantageous in fighting financial criminals and money launderers?

Customer Perceptions

AI systems enable the CDD (Customer Due Diligence) and KYC (Know Your Customer) systems, to take place at a faster rate and with greater deepness and reach. The AI-based CDD and KYC processes enable the financial institution to

Efficiently identify and collect data from a greater range of external sources which include watch lists, sanction lists, and create a factual profile of the customer.

Recognize valuable owners of customer entities by using external data faster and more efficiently.

Accumulate and reconcile customer data across internal systems to remove replication and errors and intensify the density of AML measures among customers.

Automatically enhance dubious activity reports with appropriate data from customer risk profiles or data from external sources.

Unstructured Data

There are other important steps beyond creating customer risk profiles. As a part of monitoring transactions, screening PEP, screening sanctions, and monitoring media, the AML process requires identifying and analyze the unstructured data. Every financial institution must make an effort to use the unstructured data to recognize their professional, social and political lives by inspecting a range of external sources which includes public archives, media, social networks, etc. in such circumstances, AI helps the institution to recognize those unstructured data. Once the data is collected and analyzed, AI helps the institution prioritize and categorize information to assist risk management.

Reporting Dubious Activity

AI can assist the reporting of doubtful activity by producing reports and also, by automatically filling them with accurate information. After their submission of reports to the authority, SARs goes through a process of internal reporting. AI technology can make the SAR process easy as algorithms can generate automated reports with accurate data and transmute that data into an accessible, standardized language in order to eliminate bureaucratic friction. Because of standardized language and terminology, AI increases the speed and efficiency of an institution’s AML reporting.

Noise Minimization

The AML system is complex and is a time-consuming procedure therefore it is an advantage to incorporate AI within an AML system which helps in adding speed and efficiency. But one of the major hindrances in the process is the level of noise or false positives which is the result of incomplete or inadequate data or over-sensitivity of AML steps. In such cases, AI systems play an important role by generating a significant transformative effect to the level of noise generated during the AML process. AI assists the institution to produce higher insight into customer’s transaction patterns and enables them to remove wrong and invalid alerts which makes the process costly for the institutions and inconvenient for customers. By minimizing noise, AI and machine learning tools enable AML employees to better prioritize and direct the most required money laundering alerts. By doing so AI more effectively contributes to the fight against financial crime.

Limitations of AI

In order to keep pace with the increasing risk of financial criminals and money launderers and the need to react faster to those new threats, often new AI and machine learning models are prematurely dashed into the market without proper training. This creates a huge skepticism around AI and Machine Learning technologies. Therefore, banks must remember that AI experimentation comes with diminishing returns. They should focus on performing strategic, production-ready AI micro-projects in parallel with human teams to deliver actionable insights and value.

The post AI: TECHNOLOGY TO FIGHT FINANCIAL CRIMINALS AND MONEY LAUNDERERS appeared first on Artificial Intelligence.

Three cybersecurity experts explained how artificial intelligence and machine learning can be used to evade cybersecurity defenses and make breaches faster and more efficient during a NCSA and Nasdaq cybersecurity summit.

Kevin Coleman, the executive director of the National Cyber Security Alliance, hosted the conversation as part of Usable Security: Effecting and Measuring Change in Human Behavior on Tuesday, Oct. 6.

Elham Tabassi, chief of staff information technology laboratory, National Institute of Standards and Technology, was one of the panelists in the “Artificial Intelligence and Machine Learning for Cybersecurity: The Good, the Bad, and the Ugly” session.text

“Attackers can use AI to evade detections, to hide where they can’t be found, and automatically adapt to counter measures,” Tabassi said. 

Tim Bandos, chief information security officer at Digital Guardian, said that cybersecurity will always need human minds to build strong defenses and stop attacks. 

“AI is the sidekick and security analysts and threat hunters are the superheroes,” he said. 

Here are three ways AI and ML can be used in cybersecurity attacks. 

Data poisoning

Tabassi said that bad actors sometimes target the data used to train machine learning models. Data poisoning is designed to manipulate a training dataset to control the prediction behavior of a trained model to trick the model into performing incorrectly, such as labeling spam emails as safe content. 

There are two types of data poisoning: Attacks that target a ML algorithm’s availability and attacks that target its integrity. Research suggests that a 3% training data set poisoning leads to an 11% drop in accuracy. 

With backdoor attacks, an intruder can add an input to an algorithm that the model’s designer does not know about. The attacker uses that backdoor to get the ML system to misclassify a certain string as benign when it might be carrying bad data.

Tabassi said that techniques for poisoning data can be transferred from one model to another.  

“Data is the blood and fuel for machine learning and as much attention should be paid to the data we are using to train the models as the models,” she said. “User trust is influenced by the model and the quality of the training and the data that is going into it.”

Tabassi said the industry needs standards and guidelines to ensure data quality and that NIST is working on national guidelines for trustworthy AI, including  both high-level guidelines and technical requirements to address accuracy, security, bias, privacy, and explainability.

Generative Adversarial Networks

Generative Adversarial Networks (GANs) are basically two AI systems pitted against each other—one that simulates original content and one that spots its mistakes. By competing against each other, they jointly create content convincing enough to pass for the original. 

Nvidia researchers trained a unique AI model to recreate PAC-MAN simply by observing hours of gameplay, without a game engine, as Stephanie Condon explained on ZDNet.

Bandos said that attackers are using GANs to mimic normal traffic patterns, to divert attention away from attacks, and to find and exfiltrate sensitive data quickly.

“They’re in and out within 30-40 minutes thanks to these capabilities,” he said. “Once attackers start to leverage artificial intelligence and machine learning, they can automate these tasks.”

GANs also can be used for password cracking, evading malware detection, and fooling facial recognition, as Thomas Klimek described in the paper, “Generative Adversarial Networks: What Are They and Why We Should Be Afraid.” A PassGAN system built by machine learning researchers was trained on an industry standard password list and was eventually able to guess more passwords than several other tools trained on the same dataset. In addition to generating data, GANs can create malware that can evade machine learning-based detection systems. 

Bandos said that AI algorithms used in cybersecurity have to be retrained frequently to recognize new attack methods. 

“As adversaries evolve, we have to evolve as well,” he said.

He used obfuscation as an example, such as when a piece of malware is mostly built with legitimate code. A ML algorithm would have to be able to identify the malicious code within it.

Manipulating bots

Panelist Greg Foss, senior cybersecurity strategist at VMware Carbon Black, said that if AI algorithms are making decisions, they can be manipulated to make the wrong decision.

“If attackers understand these models, they can abuse these models,” he said.

Foss described a recent attack on a cryptocurrency trading system run by bots.

“Attackers went in and figured out how bots were doing their trading and they used the bots to trick the algorithm,” he said. “This can be applied across other implementations.”

Foss added that this technique is not new but now these algorithms are making more intelligent decisions which increases the risk of making a bad one.

The post 3 ways criminals use artificial intelligence in cybersecurity attacks appeared first on Artificial Intelligence.