ABT SHIELD can instantly evaluate a request and determine if a valid user—a human—or an automated attack:

1.    ABT SHIELD Tracking Tags (https referrals) are embedded in the webpage, app or advertising assets.  Our Tracking Pixels are easily manageable by any TAG Manager system and embedded into a website. On web page load, javascript snippet loads the ABT SHIELD Tracker. This lightweight app is nonblocking and does not slow down the user experience or web application.

2.    The ABT SHIELD Tracking is now running within the user’s browser with access to network, browser and user data.

3.    Tracking data is fed to the cloud-based Bots and Trolls Detector for real-time evaluation with a Risk Score is assigned to the source of connection.

4.    Risk Score passed back to the browser or can also be passed directly to infrastructure via server-to-server integration.

5.    Infrastructure can take action based on risk score for the user’s session at app server, cloud infrastructure or network device level.



  • ABT SHIELD learns patterns and tracks bots and trolls behaviour on the basis of analysis of a high volume of internet traffic. Key publishers and online businesses are carrying out a shared traffic analysis in order to exchange experiences connected to uncovering phenomena of automated bots and trolls. Each new member of ABT SHIELD helps in building more accurate and complete SHIELD against bots and trolls by anonymously sharing signatures of threatening connections.  
  • ABT SHIELD uses advanced and scalable device tracking methodology. The technology is based on a unique method for the identification and tagging of incoming internet traffic through analysis of the deep layers of TCP/IP stack. Various activities of individual users “machines” (computers, laptops, smartphones) operating on websites are being integrated into one central “instance”/service. ABT SHIELD identifies bot or trolls-infected users’ devices regardless of the hardware and  regardless of the operating system
  • ABT SHIELD is a rocket-fast AI-based firewall. Connection sources tied to organised disinformation attacks show characteristics different from the behaviour of an average user. AI models and optimised heuristics then score in order to block or alert the session as desired. ABT SHIELD returns data in a fraction of a second, which makes it possible to take proper action.
  • ABT SHIELD API allows for building own decision models based on the ABT SHIELD mechanism.
  • ABT SHIELD provides reporting for analysed traffic and threats via a web-based tool. Attacks may easily be filtered for view by time and risk score. They can be further grouped by country, IP, user, and customer-defined parameters.
  • ABT SHIELD is designed for high scalability and frictionless integration. Integration involves placing an HTML tag referring to the ABT SHIELD system and handling the data it returns. ABT SHIELD has a neutral impact on the user experience. The system is provided as a service discrete from a customer’s own delivery infrastructure and therefore doesn’t affect site performance. It is fully compatible with CDNs and cloud services at any scale, as well as customer-managed equipment.
  • ABT SHIELD guarantees complete security of key information exchange.
ABTShield servers are currently processing millions of entries per second in search bot and trolling behaviour.