All Insights

Articles

AI Call Centre Trends 2026: What the Primary Data Shows

Ten AI call centre trends for 2026, each backed by a primary source: Gartner, Deloitte, Census, Eurostat, the EU AI Act and the FCC, with dates.

SUMMARIZE WITH:
SHARE THIS POST
ai call center trends 2026

Last updated: 9 October 2026. Every figure below links its primary source, with the date we read it (9 October 2026). Forecasts are labelled as forecasts.

AI in the call centre in 2026 is defined less by new technology than by a gap. Gartner finds 91% of service leaders under executive pressure to implement AI, yet only 24% show positive financial returns from it and only 20% have reduced agent headcount because of it. Meanwhile, rules on AI disclosure and access to a human are arriving.

This page tracks trends. For current costs, employment and pricing figures, see AI call centre statistics. For how AI fits a whole operation, see the AI call centre guide.

Jump to: Top 10 statistics · Method · Spend vs returns · Adoption · Agentic AI · Customers · Right to a human · Offshoring · Agent roles · Cost · Measured gains · Data · Practice · Chart · FAQ

Top 10 AI call centre statistics for 2026

  1. 91% of customer service and support leaders report pressure from executive leadership to implement AI in 2026. Gartner survey of 321 leaders, October 2025 (Gartner, 18 Feb 2026).

  2. 24% of service and support leaders demonstrated positive financial returns across their AI use cases. Gartner survey of 1,303 senior leaders, January–April 2026 (Gartner, 8 Jul 2026).

  3. 12%: the median share of their 2025 budget that service leaders invested in AI, the highest of the 10 business functions assessed (Gartner, 8 Jul 2026).

  4. 20% of customer service leaders have actually reduced agent staffing due to AI. Gartner, 321 leaders, October 2025 (Gartner, 2 Feb 2026).

  5. About 3x: customers are that much more likely to use third-party GenAI tools than company-provided chatbots to resolve service issues. Gartner survey of 3,566 customers, February–March 2026 (Gartner, 8 Jul 2026).

  6. 19.8% of US businesses used AI in the period ending 3 May 2026, and 37% of firms with 250 or more employees did (US Census Bureau, 26 May 2026).

  7. 20.0% of EU enterprises with 10 or more employees used AI in 2025, up from 13.5% in 2024 (Eurostat, 11 Dec 2025).

  8. $6.70 to $7.80: average cost per assisted contact, 2024 to 2026, among 720 B2C service leaders surveyed (Deloitte Digital, June 2026).

  9. 2 August 2026: the date EU AI Act Article 50 began to apply. AI systems that interact directly with people must be designed to tell them they are dealing with AI, with fines of up to €15 million or 3% of worldwide turnover (European Commission FAQ, last updated 24 Jul 2026).

  10. Forecast: more than $3. Gartner predicts the cost per resolution for generative AI will exceed $3 by 2030, "higher than many B2C offshore human agents" (Gartner, 26 Jan 2026).

How we chose these figures

Every figure on this page comes from the organisation that measured or forecast it: Gartner press releases, the Deloitte Digital and Verint survey reports, the US Census Bureau, the Federal Reserve, Eurostat, the European Commission, the Federal Communications Commission (FCC), the Quarterly Journal of Economics and the National Bureau of Economic Research (NBER). For each figure we give the sample and fieldwork dates where the source states them. All sources were read on 9 October 2026.

Three rules decided what stayed out:

  • No secondary figures. A number quoted by a blog that quotes a report is excluded unless we could read the report.

  • Forecasts are labelled. "By 2028" is a prediction, not a measurement.

  • Vendor surveys are labelled. Deloitte sells AI transformation services and Verint sells contact-centre AI. Their surveys are primary sources for what their respondents said, not neutral audits.

We also excluded market-size estimates, which vary widely by scope and are mostly published by research resellers.

What is missing: Brilo's own data. Brilo AI publishes this page and sells AI voice agents. Brilo has not yet published its own usage data, so this page carries no original Brilo figure.

Trend 1: AI spend is rising faster than proof of return

Budgets for customer service AI are growing ahead of the evidence that they pay back.

  • 91% of service leaders feel executive pressure to implement AI in 2026 (Gartner, 321 leaders, October 2025; 18 Feb 2026).

  • 12%: the median share of the 2025 budget service leaders put into AI, the highest of 10 functions. 24% showed positive financial returns across their AI use cases (Gartner, 1,303 senior leaders, January–April 2026; 8 Jul 2026).

  • Forecast: by 2028, over 50% of customer service organisations will double their technology spend "without an equivalent reduction in talent" (Gartner, 31 Mar 2026).

  • 87% of contact-centre leaders plan to increase spend on automated agent assistance. Vendor survey of 602 firms in 17 countries (Verint, State of Contact Center AI 2026).

The pattern is consistent across sources. Money and pressure are measured in the high 80s and 90s. Demonstrated returns are measured in the 20s.

Trend 2: Business AI adoption is broad but still a minority

Government statistics put AI use well below the levels claimed in vendor surveys. They measure all businesses, not contact centres specifically. No official statistic for contact-centre AI adoption was found.

  • 17% to 20% of US businesses used AI between December 2025 and May 2026, and 20% to 23% expected to within six months (US Census Bureau, Business Trends and Outlook Survey, 26 May 2026).

  • By size: 37% of firms with 250+ employees and 32% of firms with 100–249 employees, but under 20% of firms with four or fewer employees (Census, period to 3 May 2026).

  • About 18% of US firms had adopted AI by the end of 2025. Separately, 78% of the labour force works at firms that have adopted AI (Federal Reserve FEDS Note, 3 Apr 2026).

  • 20.0% of EU enterprises used AI in 2025, from 42.0% in Denmark to 5.2% in Romania. 7.2% used speech recognition and 8.8% used technology that generates written or spoken language (Eurostat, 11 Dec 2025).

In both the US and EU data, larger firms and some countries and sectors adopt far faster than the average.

Trend 3: Agentic AI is the forecast, cancellation is the risk

Agentic AI, meaning AI that takes actions rather than only answering, is where the forecasts are boldest. It is also where Gartner expects the most failures.

  • Forecast: by 2029, agentic AI will "autonomously resolve 80% of common customer service issues without human intervention, leading to a 30% reduction in operational costs" (Gartner, 5 Mar 2025).

  • Forecast: "over 40% of agentic AI projects will be canceled by the end of 2027, due to escalating costs, unclear business value or inadequate risk controls" (Gartner, 25 Jun 2025).

  • Deloitte Digital's "AI-centric" contact centres, 160 respondents that already use agentic AI, reported 85% greater profitability than low-maturity peers and were 76% more likely to report lower cost per contact (Deloitte Digital, June 2026).

Read the Deloitte figures with care. They compare mature and immature organisations, so they show correlation, not cause. Organisations that are already well run may simply adopt AI first. The two Gartner forecasts point in opposite directions, and both can come true: a few deployments reach scale while many are abandoned.

Trend 4: Customers are bypassing company chatbots

Customers are using generative AI for service, just not the chatbots companies built.

  • Customers are about 3x more likely to use third-party GenAI tools than company-provided chatbots to resolve service issues. Third-party GenAI use "nearly doubled in the past year", while company chatbot use "has remained statistically unchanged since 2022" (Gartner, 3,566 B2B and B2C customers, February–March 2026; 8 Jul 2026).

  • 58% of customers who use GenAI have used it to complete a task on their behalf, rising to 74% in B2B (same survey).

  • 54% of US customers trust human agents more than AI for product or service recommendations; 32% trust AI more (Gartner, 5,801 customers, January–February 2025; 28 Apr 2026).

  • 70% of consumers used self-service in the prior year but estimated it resolved their issue less than a third of the time. Consumers are 2x as likely to prefer the phone over self-service for simple issues, and 5.5x for complex ones (Deloitte Digital, 3,000 consumers in seven countries).

Two signals point the same way: customers want AI that completes a task, and the phone still carries the hard calls.

Trend 5: The right to a human is becoming regulation

Disclosure and access to a person are moving from good practice into law and forecast.

  • In force: EU AI Act Article 50 has applied since 2 August 2026. Systems that interact directly with people, such as chatbots and AI agents, must inform them they are interacting with AI "from the start of the first interaction", unless it is obvious. Fines reach €15 million or 3% of worldwide turnover. A grace period to 2 December 2026 covers only machine-readable marking of generated content, and only for systems on the market before 2 August 2026 (European Commission FAQ on Article 50, last updated 24 Jul 2026).

  • Forecast: by 2028, "regulatory changes related to AI will increase assisted service volume by 30%". Gartner expects customers to "request a human by default, bypassing AI agents" (Gartner, 26 Jan 2026).

For an AI voice agent, the practical response is a disclosure in the opening line and a transfer on the first request for a person. Brilo's documentation sets the transfer rule as "First ask, straight through" (When a caller asks for a person, read 9 Oct 2026). On Brilo, the opening line is set in your agent's instructions or as a fixed greeting (Write what your agent says). This is not legal advice.

Trend 6: Offshore call centres face new scrutiny in the US

In the US, the regulatory pressure on call centres is about location as well as AI.

  • On 26 March 2026 the FCC adopted a Notice of Proposed Rulemaking, FCC 26-16, on offshore call centres (FCC news release). It proposes, and seeks comment on:

  • requiring the location of the call centre to be disclosed during the interaction;

  • letting consumers transfer to a US-based centre;

  • requiring certain sensitive calls to be handled domestically;

  • requiring offshore staff to be proficient in written and spoken American Standard English.

  • The proposal covers customer service at communications providers regulated by the FCC, such as telecoms, cable and VoIP companies. It is not a final rule and does not apply to other industries (NPRM text).

Deloitte Digital flags the proceeding as a cost risk for "companies that remain heavily reliant on phone-based service". For AI planning, the point is narrower. If the rules are adopted, communications providers would face new constraints on where and how calls are handled. Communications providers should watch the proceeding before changing where customer service is handled.

Trend 7: Agent roles are being redesigned, not removed

The workforce data from 2026 shows redesign far more often than replacement.

  • 85% of service leaders are adding new tasks and responsibilities to frontline agent roles, and 75% are shifting agents into entirely new roles. Gartner, 321 leaders, September–October 2025 (28 Apr 2026).

  • 31% have implemented or are planning layoffs in response to AI through 1Q27. 63% are reducing frontline headcount gradually through attrition (same survey).

  • Forecast: by 2027, 50% of companies that attributed headcount reduction to AI will rehire staff to perform similar functions under different job titles (Gartner, 2 Feb 2026).

  • 38%: average contact-centre employee attrition in 2025 (Deloitte Digital, 720 leaders).

High attrition is why gradual reduction through attrition is the most common route: roles can be left unfilled rather than cut. For the evidence on jobs, see will AI replace call centre jobs? For planning headcount alongside automation, see AI workforce management and diallers.

Trend 8: The cost of AI is no longer assumed to fall

The old assumption was that AI gets cheaper every year while people get dearer. Gartner now forecasts the reverse for generative AI.

  • Forecast: by 2030, cost per resolution for GenAI "will exceed $3, higher than many B2C offshore human agents". Gartner cites rising data-centre costs, AI vendors pivoting from subsidised growth to profit, and more complex use cases that consume more tokens (Gartner, 26 Jan 2026).

  • Forecast: "By 2030, 10% of Fortune 500 firms will double customer service spending in order to use AI for hyperpersonalized, proactive experiences and competitive advantage" (same release).

  • Average cost per assisted contact rose from $6.70 in 2024 to $7.80 in 2026 (Deloitte Digital).

  • Leaders expect AI to cut their cost base by 28% over the next three years. That is an expectation, not a result (Deloitte Digital).

Gartner's analyst puts it plainly: "Full automation will be prohibitively expensive for most organizations." Published per-minute prices and the derived cost of a human agent minute are tracked on Brilo's AI call centre statistics page.

Trend 9: Agent assist and post-call AI are where measured gains exist

The strongest academic evidence of AI improving service work concerns AI that helps people, not AI that replaces them.

  • Brynjolfsson, Li and Raymond studied a generative-AI chat assistant that gave real-time suggestions to 5,172 customer-support agents at one firm. Issues resolved per hour rose 15% on average and 30% for less experienced and lower-skilled workers, while the most experienced and highest-skilled saw small gains in speed and small declines in quality (Quarterly Journal of Economics, 2025). The earlier 2023 NBER working paper version (w31161) also reports chats 3.8 minutes (9%) shorter, improved customer sentiment and higher retention. Caveats: text chat, not voice; a human stayed in control; one firm; 2023-era tools.

  • 44% of contact centres say AI has significantly reduced routine and repetitive agent work, and 53% say only "somewhat". 75% say AI tools have increased human agents' performance (Verint vendor survey, 602 firms).

For how agent assist works, see real-time agent assist and coaching. Automated scoring of calls is covered in AI call centre quality assurance.

Trend 10: Data and integration, not models, are the blocker

When contact-centre leaders name what holds AI back, they name plumbing.

  • Top challenges in 2026: integration of technology, systems and tools (72%), legacy systems (58%), data security and compliance (53%), cloud interoperability (36%) and digital skills gaps (33%) (Deloitte Digital, 720 leaders).

  • Only 1 in 4 contact centres has fully unified omnichannel routing, and 28% have seamlessly connected data across channels (Deloitte Digital).

  • Adoption of contact centre as a service (CCaaS) nearly tripled, from 11% in 2024 to 32% in 2026 (Deloitte Digital).

  • 58% of service leaders aim to upskill agents into knowledge-management specialists, for "accurate, continually updated content to support both AI systems and customer self-service" (Gartner, 18 Feb 2026).

A model can only answer from the knowledge and systems it is connected to. For the layers involved, see AI call centre technology explained.

What these trends mean for call centre practice in 2026

The best AI call centre practices for 2026 follow directly from the evidence above:

  1. Measure return per use case before scaling. Only 24% of service leaders show positive returns across their AI use cases. Set a baseline and a success measure for each use case before rollout.

  2. Transfer on the first request for a person. Gartner forecasts a 30% rise in assisted volume from right-to-a-human rules by 2028. Staff for it and make the transfer path work.

  3. Disclose AI in the first sentence for EU callers. Article 50 has applied since 2 August 2026.

  4. Start where gains are measured. The academic evidence of AI gains is for assisting agents (Brynjolfsson, Li and Raymond, QJE 2025). Expand to full automation call type by call type, not all at once.

  5. Fix knowledge and integration first. Integration is the top challenge for 72% of leaders. An AI agent answers only as well as its connected knowledge.

  6. Budget for AI costs that may rise. Gartner forecasts GenAI cost per resolution above $3 by 2030. Model your costs per resolved call, not per minute.

For inbound calls specifically, see AI inbound call handling.

Chart: the customer-service AI gap, 2026

Embeddable chart specification (horizontal bar chart, percentage of service leaders):

Bar

Value

Source

Sample and fieldwork

Under executive pressure to implement AI in 2026

91%

Gartner, 18 Feb 2026

321 service and support leaders, October 2025

Reduced agent headcount due to AI

20%

Gartner, 2 Feb 2026

321 service and support leaders, October 2025

Positive financial returns across AI use cases

24%

Gartner, 8 Jul 2026

1,303 senior leaders, January–April 2026

Note on the chart: the third bar comes from a different Gartner sample, and the caption must say so.

Citation line: "Source: Gartner press releases of 2 February, 18 February and 8 July 2026. Chart compiled by Brilo AI, October 2026. https://www.brilo.ai/resources/ai-call-center-trends-2026"

AI call centre trends: frequently asked questions

What are the biggest AI call centre trends in 2026? Rising AI spend ahead of proven returns (91% under pressure, 24% with positive returns), customers bypassing company chatbots, the right to a human moving into regulation, agent roles being redesigned rather than cut, and a Gartner forecast that GenAI costs will rise.

What are the best call centre AI practices? Measure return per use case, transfer on the first request for a person, disclose AI on the first turn for EU callers, start with agent assist and post-call work, fix knowledge and integration first, and model cost per resolved call.

Is AI replacing call centre agents? Not at scale so far. Only 20% of service leaders have reduced agent staffing due to AI, and 85% are adding tasks to agent roles (Gartner).

Do I have to tell callers they are talking to AI? In the EU, yes, under AI Act Article 50 since 2 August 2026, unless it is obvious. Elsewhere the rules vary. Take legal advice.

How many businesses use AI? 19.8% of US businesses (Census Bureau, period to 3 May 2026) and 20.0% of EU enterprises with 10 or more employees (Eurostat, 2025).

To test an AI voice agent on your own calls, start on the Free plan: see Brilo pricing.

All Insights

Articles

AI Call Centre Trends 2026: What the Primary Data Shows

Ten AI call centre trends for 2026, each backed by a primary source: Gartner, Deloitte, Census, Eurostat, the EU AI Act and the FCC, with dates.

SUMMARIZE WITH:
SHARE THIS POST
ai call center trends 2026

Last updated: 9 October 2026. Every figure below links its primary source, with the date we read it (9 October 2026). Forecasts are labelled as forecasts.

AI in the call centre in 2026 is defined less by new technology than by a gap. Gartner finds 91% of service leaders under executive pressure to implement AI, yet only 24% show positive financial returns from it and only 20% have reduced agent headcount because of it. Meanwhile, rules on AI disclosure and access to a human are arriving.

This page tracks trends. For current costs, employment and pricing figures, see AI call centre statistics. For how AI fits a whole operation, see the AI call centre guide.

Jump to: Top 10 statistics · Method · Spend vs returns · Adoption · Agentic AI · Customers · Right to a human · Offshoring · Agent roles · Cost · Measured gains · Data · Practice · Chart · FAQ

Top 10 AI call centre statistics for 2026

  1. 91% of customer service and support leaders report pressure from executive leadership to implement AI in 2026. Gartner survey of 321 leaders, October 2025 (Gartner, 18 Feb 2026).

  2. 24% of service and support leaders demonstrated positive financial returns across their AI use cases. Gartner survey of 1,303 senior leaders, January–April 2026 (Gartner, 8 Jul 2026).

  3. 12%: the median share of their 2025 budget that service leaders invested in AI, the highest of the 10 business functions assessed (Gartner, 8 Jul 2026).

  4. 20% of customer service leaders have actually reduced agent staffing due to AI. Gartner, 321 leaders, October 2025 (Gartner, 2 Feb 2026).

  5. About 3x: customers are that much more likely to use third-party GenAI tools than company-provided chatbots to resolve service issues. Gartner survey of 3,566 customers, February–March 2026 (Gartner, 8 Jul 2026).

  6. 19.8% of US businesses used AI in the period ending 3 May 2026, and 37% of firms with 250 or more employees did (US Census Bureau, 26 May 2026).

  7. 20.0% of EU enterprises with 10 or more employees used AI in 2025, up from 13.5% in 2024 (Eurostat, 11 Dec 2025).

  8. $6.70 to $7.80: average cost per assisted contact, 2024 to 2026, among 720 B2C service leaders surveyed (Deloitte Digital, June 2026).

  9. 2 August 2026: the date EU AI Act Article 50 began to apply. AI systems that interact directly with people must be designed to tell them they are dealing with AI, with fines of up to €15 million or 3% of worldwide turnover (European Commission FAQ, last updated 24 Jul 2026).

  10. Forecast: more than $3. Gartner predicts the cost per resolution for generative AI will exceed $3 by 2030, "higher than many B2C offshore human agents" (Gartner, 26 Jan 2026).

How we chose these figures

Every figure on this page comes from the organisation that measured or forecast it: Gartner press releases, the Deloitte Digital and Verint survey reports, the US Census Bureau, the Federal Reserve, Eurostat, the European Commission, the Federal Communications Commission (FCC), the Quarterly Journal of Economics and the National Bureau of Economic Research (NBER). For each figure we give the sample and fieldwork dates where the source states them. All sources were read on 9 October 2026.

Three rules decided what stayed out:

  • No secondary figures. A number quoted by a blog that quotes a report is excluded unless we could read the report.

  • Forecasts are labelled. "By 2028" is a prediction, not a measurement.

  • Vendor surveys are labelled. Deloitte sells AI transformation services and Verint sells contact-centre AI. Their surveys are primary sources for what their respondents said, not neutral audits.

We also excluded market-size estimates, which vary widely by scope and are mostly published by research resellers.

What is missing: Brilo's own data. Brilo AI publishes this page and sells AI voice agents. Brilo has not yet published its own usage data, so this page carries no original Brilo figure.

Trend 1: AI spend is rising faster than proof of return

Budgets for customer service AI are growing ahead of the evidence that they pay back.

  • 91% of service leaders feel executive pressure to implement AI in 2026 (Gartner, 321 leaders, October 2025; 18 Feb 2026).

  • 12%: the median share of the 2025 budget service leaders put into AI, the highest of 10 functions. 24% showed positive financial returns across their AI use cases (Gartner, 1,303 senior leaders, January–April 2026; 8 Jul 2026).

  • Forecast: by 2028, over 50% of customer service organisations will double their technology spend "without an equivalent reduction in talent" (Gartner, 31 Mar 2026).

  • 87% of contact-centre leaders plan to increase spend on automated agent assistance. Vendor survey of 602 firms in 17 countries (Verint, State of Contact Center AI 2026).

The pattern is consistent across sources. Money and pressure are measured in the high 80s and 90s. Demonstrated returns are measured in the 20s.

Trend 2: Business AI adoption is broad but still a minority

Government statistics put AI use well below the levels claimed in vendor surveys. They measure all businesses, not contact centres specifically. No official statistic for contact-centre AI adoption was found.

  • 17% to 20% of US businesses used AI between December 2025 and May 2026, and 20% to 23% expected to within six months (US Census Bureau, Business Trends and Outlook Survey, 26 May 2026).

  • By size: 37% of firms with 250+ employees and 32% of firms with 100–249 employees, but under 20% of firms with four or fewer employees (Census, period to 3 May 2026).

  • About 18% of US firms had adopted AI by the end of 2025. Separately, 78% of the labour force works at firms that have adopted AI (Federal Reserve FEDS Note, 3 Apr 2026).

  • 20.0% of EU enterprises used AI in 2025, from 42.0% in Denmark to 5.2% in Romania. 7.2% used speech recognition and 8.8% used technology that generates written or spoken language (Eurostat, 11 Dec 2025).

In both the US and EU data, larger firms and some countries and sectors adopt far faster than the average.

Trend 3: Agentic AI is the forecast, cancellation is the risk

Agentic AI, meaning AI that takes actions rather than only answering, is where the forecasts are boldest. It is also where Gartner expects the most failures.

  • Forecast: by 2029, agentic AI will "autonomously resolve 80% of common customer service issues without human intervention, leading to a 30% reduction in operational costs" (Gartner, 5 Mar 2025).

  • Forecast: "over 40% of agentic AI projects will be canceled by the end of 2027, due to escalating costs, unclear business value or inadequate risk controls" (Gartner, 25 Jun 2025).

  • Deloitte Digital's "AI-centric" contact centres, 160 respondents that already use agentic AI, reported 85% greater profitability than low-maturity peers and were 76% more likely to report lower cost per contact (Deloitte Digital, June 2026).

Read the Deloitte figures with care. They compare mature and immature organisations, so they show correlation, not cause. Organisations that are already well run may simply adopt AI first. The two Gartner forecasts point in opposite directions, and both can come true: a few deployments reach scale while many are abandoned.

Trend 4: Customers are bypassing company chatbots

Customers are using generative AI for service, just not the chatbots companies built.

  • Customers are about 3x more likely to use third-party GenAI tools than company-provided chatbots to resolve service issues. Third-party GenAI use "nearly doubled in the past year", while company chatbot use "has remained statistically unchanged since 2022" (Gartner, 3,566 B2B and B2C customers, February–March 2026; 8 Jul 2026).

  • 58% of customers who use GenAI have used it to complete a task on their behalf, rising to 74% in B2B (same survey).

  • 54% of US customers trust human agents more than AI for product or service recommendations; 32% trust AI more (Gartner, 5,801 customers, January–February 2025; 28 Apr 2026).

  • 70% of consumers used self-service in the prior year but estimated it resolved their issue less than a third of the time. Consumers are 2x as likely to prefer the phone over self-service for simple issues, and 5.5x for complex ones (Deloitte Digital, 3,000 consumers in seven countries).

Two signals point the same way: customers want AI that completes a task, and the phone still carries the hard calls.

Trend 5: The right to a human is becoming regulation

Disclosure and access to a person are moving from good practice into law and forecast.

  • In force: EU AI Act Article 50 has applied since 2 August 2026. Systems that interact directly with people, such as chatbots and AI agents, must inform them they are interacting with AI "from the start of the first interaction", unless it is obvious. Fines reach €15 million or 3% of worldwide turnover. A grace period to 2 December 2026 covers only machine-readable marking of generated content, and only for systems on the market before 2 August 2026 (European Commission FAQ on Article 50, last updated 24 Jul 2026).

  • Forecast: by 2028, "regulatory changes related to AI will increase assisted service volume by 30%". Gartner expects customers to "request a human by default, bypassing AI agents" (Gartner, 26 Jan 2026).

For an AI voice agent, the practical response is a disclosure in the opening line and a transfer on the first request for a person. Brilo's documentation sets the transfer rule as "First ask, straight through" (When a caller asks for a person, read 9 Oct 2026). On Brilo, the opening line is set in your agent's instructions or as a fixed greeting (Write what your agent says). This is not legal advice.

Trend 6: Offshore call centres face new scrutiny in the US

In the US, the regulatory pressure on call centres is about location as well as AI.

  • On 26 March 2026 the FCC adopted a Notice of Proposed Rulemaking, FCC 26-16, on offshore call centres (FCC news release). It proposes, and seeks comment on:

  • requiring the location of the call centre to be disclosed during the interaction;

  • letting consumers transfer to a US-based centre;

  • requiring certain sensitive calls to be handled domestically;

  • requiring offshore staff to be proficient in written and spoken American Standard English.

  • The proposal covers customer service at communications providers regulated by the FCC, such as telecoms, cable and VoIP companies. It is not a final rule and does not apply to other industries (NPRM text).

Deloitte Digital flags the proceeding as a cost risk for "companies that remain heavily reliant on phone-based service". For AI planning, the point is narrower. If the rules are adopted, communications providers would face new constraints on where and how calls are handled. Communications providers should watch the proceeding before changing where customer service is handled.

Trend 7: Agent roles are being redesigned, not removed

The workforce data from 2026 shows redesign far more often than replacement.

  • 85% of service leaders are adding new tasks and responsibilities to frontline agent roles, and 75% are shifting agents into entirely new roles. Gartner, 321 leaders, September–October 2025 (28 Apr 2026).

  • 31% have implemented or are planning layoffs in response to AI through 1Q27. 63% are reducing frontline headcount gradually through attrition (same survey).

  • Forecast: by 2027, 50% of companies that attributed headcount reduction to AI will rehire staff to perform similar functions under different job titles (Gartner, 2 Feb 2026).

  • 38%: average contact-centre employee attrition in 2025 (Deloitte Digital, 720 leaders).

High attrition is why gradual reduction through attrition is the most common route: roles can be left unfilled rather than cut. For the evidence on jobs, see will AI replace call centre jobs? For planning headcount alongside automation, see AI workforce management and diallers.

Trend 8: The cost of AI is no longer assumed to fall

The old assumption was that AI gets cheaper every year while people get dearer. Gartner now forecasts the reverse for generative AI.

  • Forecast: by 2030, cost per resolution for GenAI "will exceed $3, higher than many B2C offshore human agents". Gartner cites rising data-centre costs, AI vendors pivoting from subsidised growth to profit, and more complex use cases that consume more tokens (Gartner, 26 Jan 2026).

  • Forecast: "By 2030, 10% of Fortune 500 firms will double customer service spending in order to use AI for hyperpersonalized, proactive experiences and competitive advantage" (same release).

  • Average cost per assisted contact rose from $6.70 in 2024 to $7.80 in 2026 (Deloitte Digital).

  • Leaders expect AI to cut their cost base by 28% over the next three years. That is an expectation, not a result (Deloitte Digital).

Gartner's analyst puts it plainly: "Full automation will be prohibitively expensive for most organizations." Published per-minute prices and the derived cost of a human agent minute are tracked on Brilo's AI call centre statistics page.

Trend 9: Agent assist and post-call AI are where measured gains exist

The strongest academic evidence of AI improving service work concerns AI that helps people, not AI that replaces them.

  • Brynjolfsson, Li and Raymond studied a generative-AI chat assistant that gave real-time suggestions to 5,172 customer-support agents at one firm. Issues resolved per hour rose 15% on average and 30% for less experienced and lower-skilled workers, while the most experienced and highest-skilled saw small gains in speed and small declines in quality (Quarterly Journal of Economics, 2025). The earlier 2023 NBER working paper version (w31161) also reports chats 3.8 minutes (9%) shorter, improved customer sentiment and higher retention. Caveats: text chat, not voice; a human stayed in control; one firm; 2023-era tools.

  • 44% of contact centres say AI has significantly reduced routine and repetitive agent work, and 53% say only "somewhat". 75% say AI tools have increased human agents' performance (Verint vendor survey, 602 firms).

For how agent assist works, see real-time agent assist and coaching. Automated scoring of calls is covered in AI call centre quality assurance.

Trend 10: Data and integration, not models, are the blocker

When contact-centre leaders name what holds AI back, they name plumbing.

  • Top challenges in 2026: integration of technology, systems and tools (72%), legacy systems (58%), data security and compliance (53%), cloud interoperability (36%) and digital skills gaps (33%) (Deloitte Digital, 720 leaders).

  • Only 1 in 4 contact centres has fully unified omnichannel routing, and 28% have seamlessly connected data across channels (Deloitte Digital).

  • Adoption of contact centre as a service (CCaaS) nearly tripled, from 11% in 2024 to 32% in 2026 (Deloitte Digital).

  • 58% of service leaders aim to upskill agents into knowledge-management specialists, for "accurate, continually updated content to support both AI systems and customer self-service" (Gartner, 18 Feb 2026).

A model can only answer from the knowledge and systems it is connected to. For the layers involved, see AI call centre technology explained.

What these trends mean for call centre practice in 2026

The best AI call centre practices for 2026 follow directly from the evidence above:

  1. Measure return per use case before scaling. Only 24% of service leaders show positive returns across their AI use cases. Set a baseline and a success measure for each use case before rollout.

  2. Transfer on the first request for a person. Gartner forecasts a 30% rise in assisted volume from right-to-a-human rules by 2028. Staff for it and make the transfer path work.

  3. Disclose AI in the first sentence for EU callers. Article 50 has applied since 2 August 2026.

  4. Start where gains are measured. The academic evidence of AI gains is for assisting agents (Brynjolfsson, Li and Raymond, QJE 2025). Expand to full automation call type by call type, not all at once.

  5. Fix knowledge and integration first. Integration is the top challenge for 72% of leaders. An AI agent answers only as well as its connected knowledge.

  6. Budget for AI costs that may rise. Gartner forecasts GenAI cost per resolution above $3 by 2030. Model your costs per resolved call, not per minute.

For inbound calls specifically, see AI inbound call handling.

Chart: the customer-service AI gap, 2026

Embeddable chart specification (horizontal bar chart, percentage of service leaders):

Bar

Value

Source

Sample and fieldwork

Under executive pressure to implement AI in 2026

91%

Gartner, 18 Feb 2026

321 service and support leaders, October 2025

Reduced agent headcount due to AI

20%

Gartner, 2 Feb 2026

321 service and support leaders, October 2025

Positive financial returns across AI use cases

24%

Gartner, 8 Jul 2026

1,303 senior leaders, January–April 2026

Note on the chart: the third bar comes from a different Gartner sample, and the caption must say so.

Citation line: "Source: Gartner press releases of 2 February, 18 February and 8 July 2026. Chart compiled by Brilo AI, October 2026. https://www.brilo.ai/resources/ai-call-center-trends-2026"

AI call centre trends: frequently asked questions

What are the biggest AI call centre trends in 2026? Rising AI spend ahead of proven returns (91% under pressure, 24% with positive returns), customers bypassing company chatbots, the right to a human moving into regulation, agent roles being redesigned rather than cut, and a Gartner forecast that GenAI costs will rise.

What are the best call centre AI practices? Measure return per use case, transfer on the first request for a person, disclose AI on the first turn for EU callers, start with agent assist and post-call work, fix knowledge and integration first, and model cost per resolved call.

Is AI replacing call centre agents? Not at scale so far. Only 20% of service leaders have reduced agent staffing due to AI, and 85% are adding tasks to agent roles (Gartner).

Do I have to tell callers they are talking to AI? In the EU, yes, under AI Act Article 50 since 2 August 2026, unless it is obvious. Elsewhere the rules vary. Take legal advice.

How many businesses use AI? 19.8% of US businesses (Census Bureau, period to 3 May 2026) and 20.0% of EU enterprises with 10 or more employees (Eurostat, 2025).

To test an AI voice agent on your own calls, start on the Free plan: see Brilo pricing.

Automate your business with AI phone Agents

Automate your business with AI phone Agents

Automate your business with AI phone Agents

Automate your business with AI phone Agents

Call automation for healthcare, real estate, logistics, financial services & small businesses.

Call automation for healthcare, real estate, logistics, financial services & small businesses.