AI interest in customer service is widespread, but the headline numbers measure different things: plans, exploration, pilots, deployment, self-reported benefits, and customer willingness are not interchangeable. The strongest operational result in the evidence is a field study reporting an average 15% increase in issues resolved per hour with a generative AI assistant; that average concealed substantial differences among workers. Together, the findings point to real potential alongside practical limits: reliable knowledge, appropriate human handoff, and customer preference all matter.
AI customer service statistics at a glance
| Finding | What was measured | Population and timing | How to interpret it |
|---|---|---|---|
| 85% planned to explore or pilot customer-facing conversational GenAI in 2025 | Stated plans, not deployments | 187 customer service and support leaders surveyed by Gartner in July–August 2024 | A forward-looking intention reported in 2024, not a measured adoption rate for 2025. |
| 44% exploring, 11% piloting, and 5% had deployed a customer-facing GenAI voicebot | Reported status across three separate categories | The same Gartner leader survey, July–August 2024 | These are distinct stages; they should not be combined into a deployment figure. |
| 86% had implemented GenAI, initiated pilots, or started exploring it in customer service | Combined maturity stages | 1,002 executives surveyed by Capgemini Research Institute in November–December 2024 | Shows broad organizational engagement, not that 86% had implemented AI. |
| 15% average increase in issues resolved per hour | Operational productivity in a field study with access to a generative AI assistant | 5,172 customer support agents; Brynjolfsson, Li, and Raymond working paper first dated 2023 | A study-specific average, not a universal forecast; outcomes varied by agent experience. |
| 31% already realizing faster responses; 58% expected the benefit | Separate reports of realized and expected benefits | 861 executives whose organizations were exploring, piloting, or implementing GenAI for customer service, Capgemini 2025 | Expectations are not demonstrated operational results. |
| 26% already realizing enhanced satisfaction; 60% expected it | Separate reports of realized and expected benefits | The same 861-executive Capgemini survey base | These are executive reports, not a controlled causal estimate of AI’s effect on satisfaction. |
| 33% already realizing increased first-contact resolution; 52% expected it | Separate reports of realized and expected benefits | The same 861-executive Capgemini survey base | Keep reported current gains separate from anticipated gains. |
| 73% said GenAI reduced time on mundane tasks; 70% reported a reduced overall workload | Agent self-reports | Customer service agents surveyed by Capgemini Research Institute in 2025 | These are survey responses, not a measured effect for every service team. |
| 51% willing to use a GenAI assistant for customer service interactions on their behalf | Customer willingness | 4,879 customers surveyed by Gartner in January–February 2025 | Willingness is not observed use, and refers to a customer’s own assistant acting on their behalf. |
| 45% satisfied overall with service; 71% said chatbots had improved in quality over the preceding one to two years | Consumer survey responses | 9,500 consumers, summarized by Capgemini in March 2025 | Perceived chatbot improvement does not mean most consumers prefer bots for every interaction. |
What the adoption figures actually say
Gartner’s 85% figure is a plan, not an adoption count
Gartner reported that 85% of surveyed customer service leaders said they would explore or pilot customer-facing conversational GenAI in 2025. The survey was conducted in July and August 2024 and covered 187 service and support leaders. Because it asked about intended activity in a future year, the number does not establish how many organizations ultimately launched a system or how extensively they used one.
The same survey offers a useful snapshot of stages at the time: 44% said they were exploring a customer-facing GenAI voicebot, 11% said they were piloting one, and 5% said they had one deployed. These percentages describe separate reported categories for voicebots; they are not a later check on whether the 85% plan was fulfilled.
Capgemini’s 86% combines three maturity stages
Capgemini Research Institute reported that 86% of organizations had implemented GenAI, initiated pilots, or started exploring it in customer service. The executive survey was fielded in November and December 2024 and included 1,002 executives. Since the statistic deliberately combines implementation, pilots, and exploration, it is evidence of broad engagement—not an implementation-only rate.
#1 Best Overall
These two headline adoption figures answer different questions. Gartner reports leaders’ plans for customer-facing conversational GenAI and a voicebot status breakdown; Capgemini combines three organizational maturity stages for GenAI in customer service. Neither supports the simple claim that a particular share of all customer service operations had AI deployed.
What productivity and service-outcome evidence shows
A field study found a 15% average productivity gain, with uneven effects
Erik Brynjolfsson, Danielle Li, and Lindsey Raymond studied 5,172 customer support agents who had access to a generative AI assistant. Their working paper, initially dated April 2023, reported an average 15% increase in issues resolved per hour. This is an operational outcome from a particular deployment context, rather than a survey of leaders’ expectations.
The average did not describe every agent equally. Less experienced workers saw larger gains in speed and quality, while the most experienced group had small speed gains and small quality declines. That variation matters when applying the finding: an assistant may have more room to help newer workers, while experienced agents may receive less benefit or experience trade-offs. The paper’s arXiv page also includes a March 2026 manuscript version, so detailed claims should be tied to the version being discussed rather than treating the paper as a single unchanging text.
Executive reports separate benefits already realized from those expected
In Capgemini’s 2025 findings, 31% of executives at organizations exploring, piloting, or implementing GenAI said they had already started realizing faster response times, while 58% expected that benefit. For enhanced customer satisfaction, 26% said they were already seeing it and 60% expected it. For increased first-contact resolution, the corresponding figures were 33% already realizing it and 52% expecting it. These results use a base of 861 executives, distinct from Capgemini’s 1,002-executive maturity-stage finding.
The distinction between “already realizing” and “expecting” is essential. The figures report executive responses, not a controlled comparison proving that GenAI caused a change. They indicate where organizations report progress and where they anticipate gains, but do not establish the size of a causal effect on response times, satisfaction, or resolution.
Agents reported less routine work and lower workload
Capgemini reported that 73% of surveyed customer service agents said GenAI reduced the time they spent on mundane tasks, and 70% said their overall workload had fallen. These are agent self-reports. They suggest that perceived value can include reducing repetitive work, but they do not establish that all agents or teams experience the same change.
Rank #3
Customer willingness and satisfaction are different measures
Some customers are open to their own AI assistant
In a Gartner survey of 4,879 customers conducted in January and February 2025, 51% said they would be willing to use a GenAI assistant for customer service interactions on their behalf. This is willingness to use a customer-owned assistant that interacts for the customer; it is not a measure of customers accepting a company’s chatbot, nor does it show that respondents had already used such an assistant.
Gartner also cautioned that widespread adoption of customer AI assistants could create a cost risk that undermines gains from service automation. If a customer’s assistant interacts with a company’s service operation, the volume or nature of contacts could change; the willingness result alone does not quantify that effect.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePerceived chatbot improvement does not erase demand for human help
Capgemini’s March 2025 consumer-survey summary, based on 9,500 consumers, said 45% were satisfied overall with the service they received and 71% believed chatbots had improved in quality over the preceding one to two years. The same summary says virtual agents are valued for speed and convenience, while more than 70% of consumers prefer human agents for empathy and creative problem-solving.
These findings can coexist: consumers may see chatbot quality improving and value quick answers while still preferring people for emotionally sensitive or open-ended problems. The willingness figures from Gartner and the preference findings from Capgemini should not be treated as contradictory or interchangeable; they concern different populations and different kinds of AI interaction.
Knowledge quality is a readiness constraint
Gartner’s 2024 leader survey found that 61% had a backlog of knowledge articles to edit, and more than one-third reported having no formal process for revising outdated articles. This is consequential for customer-facing conversational systems and agent-assist tools that draw on company knowledge: poor or stale source material can limit the reliability of answers and recommendations.
Gartner Senior Principal Kim Hedlin summarized the issue: “Service and support leaders are eager to deploy conversational GenAI, but they cannot ignore existing issues with knowledge management.” The statistic does not prove that every system will fail when its knowledge base is weak, but it identifies an operational condition organizations reported alongside their interest in deployment.
Recommended Free Tools
How to read AI customer service claims responsibly
- Identify the stage. Exploration, a pilot, partial rollout, and broad deployment are not equivalent. A combined percentage should retain the stages it includes.
- Separate plans from outcomes. A leader’s intention to pilot a system in a future year is not evidence that the pilot occurred.
- Separate realized reports from expectations. “Already realizing” and “expect” are different answers, even when shown side by side.
- Check who answered. Executives, service agents, and consumers report different perspectives. A response from one group should not be attributed to another.
- Check the outcome and method. Issues resolved per hour, response time, first-contact resolution, customer satisfaction, and workload are distinct measures. An operational field-study result, survey self-report, and forecast do not offer the same kind of evidence.
- Keep AI use cases distinct. Agent assistance, a company’s customer-facing chatbot or voicebot, and a customer’s personal assistant acting on their behalf are different applications.
- Account for task and worker differences. The field-study average concealed variation by experience; a result for one work setting does not guarantee the same effects on complex cases or experienced teams.
- Include knowledge and human escalation. Content maintenance and a way to reach a person matter, especially where an issue requires empathy or creative problem solving.
Sources and survey scope
- Gartner, “Gartner Survey Reveals 85% of Customer Service Leaders Will Explore or Pilot Customer-Facing Conversational GenAI in 2025,” December 9, 2024. The reported survey covered 187 customer service and support leaders in July–August 2024.
- Capgemini Research Institute, Unleashing the value of customer service: The transformative impact of Gen AI and Agentic AI, 2025. The reported fieldwork took place in November–December 2024; executive, agent, and consumer findings use different survey bases.
- Capgemini, “Generative and agentic AI set to transform customer service into a strategic value driver for businesses,” March 13, 2025. Its methodology summary includes 9,500 consumers, 506 agents and supervisors, and 1,002 executives.
- Erik Brynjolfsson, Danielle Li, and Lindsey Raymond, “Generative AI at Work,” working paper initially dated April 2023; the arXiv page also exposes a March 2026 version.
- Gartner, “Gartner Identifies Three Trends That Will Shape The Future of Customer Service,” June 25, 2025. The customer survey covered 4,879 respondents in January–February 2025.
Frequently Asked Questions
Does the 86% figure mean that 86% of companies had deployed AI in customer service?
No. Capgemini’s figure combines organizations that had implemented GenAI, started pilots, or begun exploring it; it is not an implementation-only percentage.
Is the 15% increase in issues resolved per hour a guaranteed result?
No. It is the average reported in one field study of 5,172 customer support agents, and the study found different effects by worker experience.
Do the surveys show that customers prefer chatbots to people?
They do not establish that broad preference. Capgemini reported perceived chatbot improvement and also reported that more than 70% preferred humans for empathy and creative problem-solving.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Free tools Windows power users keep installed
One-click scans. No signup required.
