No Fortune 500 Going Agentless by 2028: Gartner Backs Human-AI Balance in Customer Service

AI Won't Replace Customer Service-It Will Refocus It

There's a gap between headlines and reality. A recent report from Gartner says a fully agentless future is unlikely and, frankly, undesirable for customer service teams.

The firm states that none of the Fortune 500 will remove humans entirely from service roles by 2028. The message: use AI to shift work, not erase it.

What the data says

By 2027, about half of the organizations that tried to replace their service workforce with AI will drop those plans. Many won't meet the staffing cuts they forecasted.

As Kathy Ross, Senior Director Analyst for Gartner's Customer Service & Support practice, put it: "We expect fewer human agents, but not completely agentless organizations." The winners will balance technology with the human touch and redeploy teams toward growth and customer satisfaction.

Why humans still matter

AI handles routine, repetitive, and well-defined problems. It struggles with exceptions, high-risk scenarios, and emotionally charged moments where trust is on the line.

We've already seen the cost of going too far. Swedish fintech Klarna reportedly rehired humans after a wave of complaints, and Australia's Commonwealth Bank reversed cuts when an AI voice bot drove call volume up. Over-automation can boomerang fast.

What this means for support leaders

Use AI to scale your people, not replace them. Think deflection and acceleration, not elimination.

• Automate the simple: identity checks, status queries, order lookups, password resets.
• Assist the complex: draft replies, summarize threads, surface knowledge, suggest next steps.
• Reserve humans for edge cases, sensitive issues, and moments that build loyalty.

A practical playbook to blend AI + human support

• Map intents: List top contact reasons. Mark which are automatable. Define clear failover to humans.
• Set guardrails: Escalate on account risk, vulnerable customers, repeated failures, or sentiment dips.
• Start small: Run AI in shadow mode. A/B test flows. Expand only after quality hits target.
• Upgrade knowledge: Keep your KB clean. Add structured answers. Build feedback loops from agents.
• Measure what matters: Containment rate, FCR, CSAT by channel, escalation quality, handoff friction.
• Invest in roles: Conversation design, bot operations, quality assurance, and AI ops ownership.
• Compliance and trust: Review high-risk flows. Log decisions. Protect data. Make the "talk to a human" path obvious.
• ROI with context: Track cost-to-serve and deflection, but also churn, NPS, and revenue from saves and expansions.

For frontline agents: upgrade your craft

• Use AI to pre-read the case: summary, tone, and key events. Then personalize your response.
• Confirm understanding in one sentence. Offer one clear next step. Avoid jargon.
• Tag edge cases and broken flows. That data trains both the bot and the business.
• Double down on empathy, negotiation, and problem framing. These skills compound.

Team enablement in 90 days

• Days 0-30: Audit top journeys. Ship two low-risk automations. Train agents on AI-assisted workflows.
• Days 31-60: Add escalation rules. Launch quality reviews for AI-assisted responses. Publish a playbook.
• Days 61-90: Expand to two more intents. Report ROI and CX impact. Reassign saved capacity to retention and high-value accounts.

Bottom line

AI should handle the simple. Humans should own the complex, high-risk, and relational. That balance is the strategy.

If you lead support, your edge isn't fewer people-it's smarter orchestration of people and systems.

Further resources

• Gartner Customer Service & Support insights
• Practical AI training by job function to level up your team's skills