Article • 4 min read
Personal AI is next in line to disrupt CX - here's how to prepare
AI is no longer just a tool for businesses—customers are deploying it too, using personal assistants to interact with brands and handle tasks on their behalf.
출처 Jason Maynard, Chief Technology Officer (AMER & APAC), Zendesk
최종 업데이트: February 26, 2025
AI is poised to further transform customer experiences, but this time it’s end users who’ll be deploying it. Need to book that last-minute dinner reservation? Your personal AI assistant can do it. What about initiating a return for that item you just couldn’t live without, but still haven’t opened? Sure, it can help there, too.
A personal AI assistant that can take all those pesky, to-dos off your plate may sound like a grand proposition, but the reality is that this will all feel quite normal in just a few years time. If you’re in need of support, you won’t have to lift a finger—personal AI will handle it and in a fraction of the time.
AI assistance is already integrated into tasks like drafting emails, organizing photos, and answering questions, but now we’re at a point where it can interact with businesses on our behalf. People are already using assistants that can call a restaurant staff to book, modify, or cancel a reservation all without ever having to pick up the phone. And as more and more customers turn to AI Agents for things like checking on an order status, getting a refund, or canceling a subscription, businesses will need to adapt—rethinking scale, channels, and integrations to keep up with this shift.
The easier it is to connect with customer support, the more likely it is that people will do it. We expect customer engagement to rise some three to five times its current rate in the next five years, with AI assistants becoming a major channel for B2C brands. But as these engagements become more complex—such as updating an account or processing a return—the interactions will become more streamlined, direct, and instantaneous. AI Agents will talk directly to business systems via APIs, completing tasks in mere milliseconds and giving businesses the chance to standardize new protocols in making AI-to-AI interactions more efficient.
Despite the potential, there are still significant barriers that will stop many companies in their tracks. Siloed backend systems, API and workflow integration, knowledge management, and privacy and security concerns are just a few. As we push closer to this next iteration of AI-powered service, here are three steps that businesses can take now to set themselves up for future success:
1. Review how LLMs answer common customer questions
I’m sure you’ve heard of SEO optimization, but now you must also consider how your content is optimized for LLMs. Unlike SEO, which targets keywords like “best running shoes,” you should be optimizing for more complex and conversational queries like “what running shoes are best for icy conditions and why.” Focus your content on being factual, concise, and authoritative, making it more likely to appear as a cited source in LLM summaries. That said, LLMs still bias toward using trusted sources and domain authority, so establishing high-quality content that is referenced with backlinks is still a sure-fire way to ensure your content gets picked up.
2. See how easy AI agents can navigate your website or product UI/UX
AI agents have advanced remarkably in their ability to navigate websites and systems on a user’s behalf. Agent tools like Anthropic’s Computer Use and OpenAI’s Operator, for instance, rely on APIs, structured data, and interfaces to execute tasks such as purchasing, booking, or gathering personalized information. They work best when data is structured and tagged properly, using things like schema.org and WCAG standards to aid with identifying product details, pricing, availability and navigating actions like cancellations or status updates.
3. Establish a foundation for AI Assistant workflows
Zendesk Labs is building a platform to enable seamless AI-to-AI interactions between personal assistants and brands. Today, Zendesk customers can create AI Actions for Agent Copilots, but soon, these will also be available to consumer AI assistants. This will allow AI to handle end-to-end requests, like checking a flight or making a spa booking, but connecting directly to backend systems.
It’s early days, but the progression from basic assistance to full AI-to-AI engagement between customers and companies is an all but certain reality. Companies that move quickly to embrace early adoption and refine their AI strategies will ultimately shape the future of customer experience—where AI personal assistants don’t just improve interactions, but revolutionize them.
