How Leading Vendors Innovate While Earning The Trust Of Customer Service Teams

More than 650 conversational AI vendors compete for relevance in a market shaped by the needs of the roughly 15 million contact center agents working globally today. Conversational AI solutions perform rote tasks that no human really wants to perform: Think answering basic questions, canceling credit cards, or opening service tickets. The 14 vendors included in The Forrester Wave™: Conversational AI Platforms For Customer Service, Q2 2026 excel at using modern AI (large language models, generative AI, and agentic frameworks) to do this routine work for agents, thereby improving customer experiences while saving organizations money.

Conversational AI platforms must walk an interesting tightrope. They need to evolve and incorporate successive waves of AI innovation that expand the potential of their platforms by an order of magnitude almost annually. In order for an organization to bring an AI–powered customer self-service application into production, the platform must pass muster with risk–averse legal and compliance teams. This requires taming AI with tools for enterprise-grade observability and enforceable guardrails. In addition, conversational AI platforms must fit into real-world contact center environments: integrating with entrenched, long-lived IT systems and connecting to telephony and contact-center-as-a-service platforms.

The 14 vendors in this Wave successfully walk that line. What makes them stand out? One, it’s technological advancements to keep up with the “Wild West” that is AI technology. Two, it’s a commitment to understanding the very specific demands of customer service and contact center organizations. Some key traits that make these vendors stand out in this market include:

  • Understanding how to fit into the existing customer service IT environment. Organizations want to use conversational AI systems to perform transactions for their customers and keep their customers’ data safe. These systems support modern connectivity standards, primarily Model Context Protocol, to connect to modern AI systems. Additionally, they need RESTful APIs and software development kits to connect to CRM and other back-end systems. Plus, they must be able to easily to connect to telephony networks and messaging applications to meet customers where they are.
  • The ability to protect customer data and minimize hallucinations. The vendors in this Wave support a variety of customer data security standards (even FedRAMP in a couple of cases) to help ensure data protection. Some provide on premises options for data storage or can meet data sovereignty rules. All these vendors have a variety of contractual limitations to ensure the proper use of customer data. These systems are not “black boxes” that just run without an audit trail — these vendors provide powerful tools that provide visibility into how the AI is thinking, what it did, and why it made each choice. These capabilities help organizations troubleshoot their systems, minimize hallucinations, and keep things running in order.
  • A mix of practical and promising development tools. Customer service teams generally have minimal developer support to build their self-service applications. In response, vendors in the space offer no-code development tools. With new AI capabilities moving quickly, customer service teams now need both programmer support and no-code tooling. Most of the vendors in this Wave provide a set of practical pro-code and no-code tools in an environment that empowers developers to work as a team. Meanwhile, exciting new capabilities related to AI–driven coding, discovery, and testing tools are available from many of these vendors. These early tools demonstrate the power to deliver a new level of automated application creation today that we see will become mainstream before long.
  • Real agentic frameworks. Currently, there is zero appetite in this market for fully autonomous agentic applications. In fact, many organizations struggle to get simple generative AI–driven FAQ applications into production. This reality is not stopping leading conversational AI vendors from reinventing their products as agentic frameworks. The agentic approach to application delivery lends itself to building modular and efficient applications. Most vendors allow a mix of deterministic and predictive application approaches to allow organizations to give bots as much or as little autonomy as they are comfortable with, knowing that the framework to increase autonomy is in place when they need it.

Forrester clients, please book a guidance session or inquiry with me to learn more about these findings — and what they mean for your organization.