Enterprises will be looking at ways that AI tools can be used to protect their business from threats, comply with regulations and safeguard data. Training of AI systems will be an especially critical factor in ensuring adherence to industry regulations.
This blog is part of our ongoing series, IPsoft’s 2019 AI Trends, detailing what we believe will be the dominant developments and movements in the Enterprise AI market next year.
For many people, the thought of AI listening to their conversations or inappropriately using our information fills us with fear. We’ve heard tales of Alexa recording private conversations and sending them to recipients without permission or chatbots going “rogue” and posting unsuitable messages. This leaves many of us with a concerned outlook for our future personal security in an AI-driven world.
Enterprises are similarly concerned with protecting customer data and their own IT systems when using AI. In a recent survey by HfS Research, almost 50% of leaders stated security concerns as a large obstacle to introducing cognitive solutions within their IT department. However, the ways that AI can provide security benefits are extensive and will take on new importance as AI adoption accelerates in 2019.
Enterprises Will Use AI for Advanced Security
In 2019, IT security will continue to be a priority, particularly as new data regulations such as GDPR emerge and companies face potential lawsuits for failing to protect customers’ data. Enterprises will be looking at ways that AI tools can be used to protect their business from threats, comply with regulations and safeguard data. Training of AI systems will be an especially critical factor in ensuring adherence to industry regulations.
Let’s review what will be some of the top AI security trends for next year.
Enterprises will be implementing cognitive agents and intelligent biometric systems as first points of contact when a customer reaches out for customer service or sales support. Instead of an Interactive Voice Response (IVR) system or a human agent asking a number of repetitive questions to authenticate a customer’s identity, AI will make this process both faster and more accurate. Cognitive agents can verify a user, backed by real-time analytics, and voice or fingerprint biometric systems. Adding an AI layer enables enterprises to quickly detect when patterns change and reduces human error. As noted by analyst firm Opus Research, next year “practical-thinking enterprises will embed intelligent authentication to help prevent fraudsters from gaining access to valuable data and increase customer loyalty with pleasant, seamless customer interactions.”
A leading gaming company is currently using Amelia, IPsoft’s cognitive agent, to help stop phishing attacks during customer service calls. Amelia accurately detects and stops impostors who attempt to access the accounts of genuine players. These scams are frequently directed at live chat agents who provide player support. Amelia was able to verify customers faster and more accurately, and human agents now spend more time on high-value customer service calls.
Adherence to Regulations
Cognitive agents are trained to follow specific rules and processes and can quickly change behaviors to comply with new regulations. Employees are sometimes resistant to change, and training multiple agents on a new law or process takes time. In addition, agents understand that they are measured on customer satisfaction and want to appease unhappy customers, occasionally providing offers or discounts that they may not be fully authorized to execute in order to receive positive customer feedback and ratings. Cognitive agents handle customer service by following the processes on which they have been trained, reducing human error and business risk — not just by preventing the unauthorized sharing of customer data, but by also lowering the potential for human agents to offer customer giveaways when they shouldn’t. Although contact centers do need some flexibility in handling customer service requests to ensure high satisfaction and customer loyalty, AI can work within set parameters, handing over high-value tasks to human agents. Cognitive agents abide by security rules to ensure that sensitive data is not shared or stored during conversations with agents or even internally within an organization.
When used as a whisper agent, the cognitive agent can also help employees find information faster. For example, in the heavily regulated insurance industry, the whisper agent can alert a human colleagues to the latest laws during a customer call, rather than relying on the employee to remember every law and subsequent updates.
Predicting and Preventing Cyber Attacks
Through a combination of real-time analytics and AI, enterprises can prevent cyber-attacks by using AI solutions to monitor IT systems for traffic spikes or anomalies. Humans are still required in the process for investigating and verifying attacks, but by using software to alert them to potential problems before they escalate, the enterprise can resolve problems sooner. In addition, with self-learning capabilities AI will spot similar attacks faster over time as it gains experience with various types of security breaches.
IT access management
New and departing employees need to be granted IT systems access or have access revoked quickly in order to optimize the staffing process. Managing access manually for each individual application is time consuming and tedious, and IT staff typically do not view this as a high-priority task. However, the timeliness of IT access impacts staff productivity as well as posing a security risk to the business when staff leave and retain access to company data long after they should. With intelligent process automation, this procedure can be managed quickly and securely, and as a result, AI will see new use cases emerge next year in HR that will transform and secure employee onboarding and offboarding.
A global telecom provider implemented IPsoft’s AIOps (formerly 1Desk) to offboard staff faster. By using the platform to handle 2,400 access restriction requests for 200 workers each month, the company was able to reduce the offboarding process from hours to fifteen minutes. Resolution times were reduced by more than 80% and the company ensured that access management was carried out 24/7/365 to reduce delays, errors and security risks.
Ethics and Customer Data Protection
As AI investments continue in 2019, there will be increased conversations around protecting customer and enterprise data. Enterprises and vendors will collaborate to reduce bias from algorithms and ensure that new AI solutions are benefiting everyone through fair and ethical methods. Companies should consider implementation plans that include strict guidelines for use and storage of customer data as well as frequent training for both staff and the AI solution on rules and regulations.
While AI can bring new complexity and change to the organizational structure, enterprises will see the value it provides in predicting issues, authenticating customers and securing systems.