Bot Detection & Protection | Adaptive Response

Advanced Bot Detection, Adaptive Response

Bots are evolving and it’s become increasingly difficult to detect and block them with traditional solutions, such as perimeter security, web application firewalls, and content delivery networks. Kount’s adaptive bot protection is the next generation, event-based solution that combines network, device, and behavior signals with billions of identity trust signals from the Identity Trust Global Network™ to accurately identify and classify even the most sophisticated bots, enabling you to quickly adapt your defenses and protect the complete customer journey.

Malicious Bot Activity

Identify and Block
Malicious Bot Activity

Good Bot Activity

Recognize and Remove
Friction for Good Bots

Questionable Bot Activity

Classify and Adapt Responses
for Questionable Bots

Bot Protection Across the Entire Customer Journey

Kount’s unique view of the complete customer journey allows it to establish a baseline of normal behavior, and then to quickly identify abnormal, high risk behavior. By expanding bot prevention and detection from checkout to earlier points in the customer workflow, Kount offers an advantage that is critical to mitigating the cost of bot attacks on business and customers, as bots can attack multiple points in the digital journey.

Account
Creation

Account
Login

Loyalty
Redemption

Coupon
Redemption

Gift Card
Redemption

Add to
Basket

Make
Payment

Request a demo to see how it works

Detect, Classify, and Stop Malicious Bots

Existing bot protection solutions typically protect the perimeter and identify bots using network characteristics, device characteristics, and behavior characteristics. Kount’s event-based bot protection applies a layered approach to identify and segment bots, examining these characteristics along with past behaviors and identity trust signals to better understand bot behaviors and the identity behind the interaction. When a malicious bot is identified, the decision data of the interaction is fed back into the Identity Trust Global Network so that other businesses, who would otherwise be unaware of an attack, can prevent similar attacks. Using advanced reporting, in-depth insights into customer behaviors and bot trends can be identified and used to inform future policies and strategies.

Network-Level Detection

Bot detection based on network characteristics works by analyzing network traffic flow during constant time intervals. But sophisticated bot attacks can circumvent aggregate detection by using a “slow-and-low” approach that distributes the attack over multiple IPs and over a long period of time. This attack is normally under the thresholds that would trigger an alert. To detect sophisticated bot attacks, Kount looks more deeply at the activity associated with IP addresses, including IP use and geographic distribution, and increases in individual event anomalies. By assessing the details of the user’s behavior and signals from the Identity Trust Global Network, Kount can uncover patterns that suggest malicious intent.

Device-Level Detection

Device-based bot detection stops attacks that mimic device attributes to bypass security measures. Basic attacks simulate regular network traffic, but more sophisticated attacks can match a user’s device exactly, either because the device is controlled by a real human or a user recorded action. To stop these attacks, Kount’s Device Intelligence allows businesses to collect detailed device information from site visitors, such as device ID, device type, geolocation, and more, to reveal a complete view of the device. Device Intelligence accurately identifies a device and flags anomalies so businesses can deliver personalized experiences to known customers, and detect and stop bots in real time.

Behavioral-Level Detection

Sophisticated bots use advanced AI to respond in ways similar to a real user. Kount’s advanced approach can detect sophisticated bots by looking beyond the behavior, networking the interaction to Kount’s extensive payments, account, and identity data. Kount looks beyond the firewall to connect the bot to legitimate and trustworthy interactions, or to suspicious events and fraudulent activity. With this intelligence, businesses can recognize the identities behind behaviors in real time to improve bot detection and business outcomes.

Identity Trust Global Network™

Kount has the unique ability to network multiple data types across organizations, industries, and geographies in order to identify patterns, determine levels of trust, and uncover anomalies. When the same device, IP, email, payment token, or mailing address is used somewhere on the Kount network, Kount is able to assess the risk in real time in the context of the attack. By linking network, device and behavioral characteristics to trust and risk signals from the Identity Trust Global Network, Kount quickly delivers a more accurate assessment of risk in every interaction to effectively stop fraud without adding unwanted friction.

Advanced AI and Machine Learning

Kount evaluates every individual event, from account creation, account login, card verification, loyalty or coupon redemption, to payment, to identify bot-like behavior and mitigate risk. Kount identifies high risk behavior in real-time using unsupervised and supervised machine learning, identifying and linking millions of unique events within 300ms. And by using both supervised and unsupervised machine learning, Kount can learn from past decisions to prevent known attacks and detect and stop emerging attacks.

Schedule a Demo to See Kount’s AI-Driven Fraud Prevention in Action

Classifying Bots

Determining if a bot is good or malicious may depend on the perspective of a department or business. Among departments, a bot that appears harmless to cyber security can disrupt ecommerce or customer experience. And among organizations, one may allow bots for search engine optimization, automated purchases, and quotes, while another blocks similar bots. Looking deeper, one retailer may choose to block bots that purchase large quantities of a particular product at a reduced price, as this can signal a retail arbitrage attack. Another retailer may choose to allow the bot to run, as their customers are authorized resellers and often buy larger quantities of discounted goods.

With Kount’s solution, businesses can customize their response to bots in real-time based on the bot’s behavior and the desired business outcomes.

Good Bot Activity

Search Engine Bots
Copyright Bots
Site Monitoring Bots
Commercial Bots
Feed Bots

Questionable Bot Activity

Social Media Bots
Scraper Bots
Quoting Bots
Ticketing Bots
Automated Purchasing Bots

Malicious Bot Activity

Card Testing / Card Cracking Bots
Account Testing Bots
DDoS Bots
Brute Force Bots
Spam Bots
Denial of Inventory Bots
Ad Fraud Bots

Adaptive Response

Kount gives fraud teams the power to manage and mitigate bot attacks in real-time: they can approve an interaction, block it, or prompt for step-up authentication. Businesses are empowered to make quick strategic decisions, easily adjusting risk policies to respond to changing fraud conditions without expending time or revenue to develop new code or submit change requests. Fraud teams can become more agile, responding quickly to make policy adjustments in real time based on bot behaviors.

Advanced Reporting

Data and graphics within Kount’s solution help businesses understand how bots are affecting their business. Rich data about each event provide in-depth insights into customer behavior and bot trends. Custom reports can reveal hard-to-detect fraud by showing comparisons over time of transaction volumes, decline rates, new email addresses, numbers of payment cards, IP locations, order details, and more – all of which can be used to identify bot behavior.

Real-Time Protection

Kount focuses on protecting key interaction points across the customer journey, while reducing the false positives and unnecessary friction that affect solutions with limited data and a lack of real-time AI. Built on the Identity Trust Global Network, Kount’s event-based bot detection is able to quickly distinguish between good, questionable, and malicious bots, enabling a fine-tuned approach to bot protection and mitigation throughout the digital journey.

Event-Based Bot Detection