ChatGPT is becoming your new CX operating system
Plus: Plus: Google’s “CC” could turn the daily briefing into an AI habit
📅 December 22, 2025 | ⏱️ 4-min read
Good Morning!
The Executive Hook:
If you’ve ever rolled your eyes at “brand voice guidelines” sitting in a PDF while your AI says something… technically correct but emotionally wrong — today’s theme is for you. The AI era isn’t just about automation. It’s about control: control over tone, control over integrations, and control over what customers feel when they interact with your systems.
🧠 THE DEEP DIVE: ChatGPT’s App Directory — the “tools menu” becomes your CX stack
The Big Picture: OpenAI added an in-product app directory in ChatGPT so people can browse and attach approved apps directly inside a conversation.
What’s happening:
ChatGPT now has an app directory where users can browse and add approved apps (including what used to be called “connectors”).
OpenAI is positioning “apps” as both interactive experiences and secure connections that let ChatGPT reference third-party tools and data inside the chat.
The implication is big: instead of customers bouncing between channels and portals, the conversation becomes the place where work gets done.
Why it matters:
For CX leaders, this is the quiet shift from “chatbot on the website” to “chat as a workspace.” If your support, commerce, and knowledge flows can be pulled into one conversation, then the experience is less about channel strategy and more about orchestration: what data is available, which actions are allowed, and how you prevent a “helpful” AI from doing the wrong thing at speed.
The takeaway:
If you’re evaluating AI for customer-facing use, start asking a different question: “What do we want AI to do inside the conversation — and what must it never do?” Then build your app/integration strategy around that boundary.
Source: TechCrunch
📊 CX BY THE NUMBERS: AI is taking more of the queue — fast
Data Source: Salesforce, 7th State of Service (survey of 6,500 service professionals; conducted Apr 25–Jun 6, 2025; published Nov 13, 2025)
50% of service cases are expected to be handled by AI by 2027 (up from 30% today).
AI jumped from #10 to #2 on service leaders’ priority list in one year (with improving CX still #1).
Service reps using AI report 20% less time on routine cases — about 4 hours per week freed up for more complex work.
The Insight:
This isn’t just “more automation.” It’s a capacity shift. If AI takes a bigger share of routine work, your CX edge moves to what happens next: escalation design, exception handling, and the quality of the human moments that remain.
Source: Salesforce
🧰 THE AI TOOLBOX: Google “CC” — a morning briefing agent for your inbox
The Tool: Google’s experimental AI assistant “CC” sends a daily “Your Day Ahead” email briefing using context from Workspace.
What it does:
It summarizes your schedule and priorities, pulls in relevant context from Gmail/Calendar/Drive, and nudges you toward next actions — without you having to ask.
CX Use Case:
Frontline leadership briefings: Auto-summarize what leaders need to know each morning (escalations, outages, staffing gaps, top customer themes).
Agent supervisor rhythm: A daily digest that turns yesterday’s contact center “noise” into today’s focus (top drivers, repeat contacts, process breaks).
Trust:
Proactive AI is powerful — and risky — because it shapes attention. If you deploy this internally, make sure the briefing shows why it prioritized something (and what it might be missing). That transparency is the difference between “helpful” and “creepy.”
Source: The Verge
⚡ SPEED ROUND: Quick Hits
FTC probes Instacart over an AI-driven pricing tool — dynamic pricing without transparency is a trust-killer, especially when customers feel singled out.
Source: ReutersOpenAI rolls back its model “router” default for many users — a reminder that “smart” isn’t enough; consistency and predictability are core experience features.
Source: WIREDFortune flags how “AI-driven layoffs” are getting framed (and misunderstood) — for CX orgs, this lands as employee anxiety and customer skepticism at the same time.
Source: Fortune
📡 THE SIGNAL: When AI gets good, your policies become the product
AI is hitting that point where it can sound right most of the time — and that’s when CX actually gets tougher. Not because the tech is breaking, but because it’s confidently moving through conversations without the kind of judgment a seasoned agent uses automatically.
So the advantage shifts to the stuff nobody gets excited about in a steering committee: what data the AI can pull, what it’s allowed to change, when it has to stop and ask a human, and how it explains itself in plain English.
Because customers don’t experience “your model.” They experience your rules — and the few moments where those rules either feel fair and helpful… or they don’t.
If you want a real CX edge in 2026, put energy into the parts customers can’t name but always feel: clear boundaries, simple explanations, and handoffs that don’t make people repeat themselves when things get nuanced.
See you tomorrow!
👥 Share This Issue
If this issue sharpened your thinking about AI in CX, share it with a colleague in customer service, digital operations, or transformation. Alignment builds advantage.
📬 Feedback & Ideas
What’s the biggest AI friction point inside your CX organisation right now? Reply in one sentence — I’ll pull real-world examples into future issues.