No Jitter kicked off 2026 with anoutlook on several trendslikely to lend greater impetus to the colliding galaxies of agentic AI, contact centers and unified communications and collaboration. As we close out the first two-thirds of 2026, let’s check in on where those trends are.
AI agents will augment rather than replace people
As more enterprises move from pilot project to full implementation, more leaders embrace an AI model centered around the technology complementing, not replacing, human workers. For example, arecent Talkdesk surveyfound that nearly all CX leaders believe ahybrid workforce of AI and peopledelivers beneficial value and nearly one in five organizations view AI agents more as labor than technology.
If AI agents are “digital labor” — a term that has itself, generally, become disused — then the natural next question is:Who pays for them? The cost of agentic AI became a flashpoint as many companies capped per-user token usage to prevent costs from spiraling out of control. Vendors are themselves experimenting with shifting from simpler per-seat pricing to consumption-based or a hybrid model that seeks toprovide budget stabilityfor customers while encouraging use of generative- and agentic AI-powered capabilities.
Despite the augmentation narrative, many vendors and analysts alike acknowledge that, indeed, some jobs will be lost. For example, Forrester forecasted a 13.5% decline incontact center agent jobsover the next five years as AI-powered voice and text automation handles more customer service issues.
Potential job or role reductions among other types of knowledge workers are as hard to estimate as the ROI gains from AI are to quantify. All these potential gains still require humans in the loop, training for the human users, as well as observability and strict governance of what the AI agents are doing because agentic AI remains non-deterministic and potentially the bestposter child for Murphy’s Law.
The year began with Salesforce galumphing into the CCaaS space withAgentforce Contact Center. In May, Zendesk announced the general availability of CCaaS as a natively integrated component of theZendesk Resolution Platform. Other vendors, likeKustomerandCreatio, have already blurred the lines between systems of record and systems of engagement. Still other vendors, like ServiceNow are simultaneouslyencroaching on CRMwhile maintaining strong partnerships with CCaaS vendors likeGenesys,Five9,NiCE,ZoomandVonage.
All “pure” contact center vendors maintain long-standing integrations tosystems of record. Contact centers handle the routing, telephony and interaction, while the customer data is all stored in various types ofsystems of record.
“AI is making engagement more valuable, even a newoteTalkingPointz Lead Analyst Dave Michels. “The CRM providers want to add engagement data to their repository, and the CCaaS vendors are realizing how AI makes database solutions trivial.”
AI makes it much easier to derive, update and maintain customer data — and then use it to extract customer insights that could lead to improvements across the entirety of the customer journey and experience. The end goal is better service, loyalty, retention and revenue over time.
Human and AI agents alike needdata and contextto deliver great customer service. Customersdon’t want to repeatthemselves. Desk workers would prefer their AI helperproduces relevant, correct responsesrather thanoutputs they can’t trust.
Incorporating context into CX might involve knowing who the customer is as they go through call triage, looking up their recent orders, divining why they might be calling and arming either type of agent with that context before directly interacting with the customer. Similarly, an AI agentassisting a knowledge workermay function better if it can access their human counterpart’s calendar, email, personal and relevant corporate documents, threads, meeting transcripts, action items, etc.
That’s the theory, at least.
In reality, delivering on this vision involves meticulous integrations, proper permissions and identity management, orchestration and coordination across back-end systems, observability and governance tools, and possibly financial controls to make sure the human/agent doesn’t spend hundreds of dollars on tokens versus the pittance it costs to use conventional means.
Training peoplehow to use AI can help them get better results from the agentic systems they’re increasingly being required to use. Yet it’s not just about how to use AI systems, it’s aboutcritically evaluatingwhat the tools produce.
Why check in with how our 2026 is holding up? Because it’s important to show how dynamic and ongoing these trends are. And to let you know that sometimes, following these stories takes us in unexpected directions.
Beginning September 8, I’ll be moving over toCX Diveas a reporter covering customer experience initiatives and businesses’ relationships with their customers. I’ve looked a lot at the technology-focused “here’s what is coming to the market” aspects of CX technology in my No Jitter coverage; over at CX Dive, I’ll get to explore the business-driven whys and customer outcomes.
Follow me over to CX Dive — and thank you for reading me on No Jitter.
