<think>
Let me analyze the search results for news about AI agents and agentic AI in customer service from October 6-14, 2025.

Key news items related to AI agents/agentic AI:

1. **Qualtrics Report (Oct 7, 2025)** - CX quality improving but AI customer support is failing. Nearly 1 in 5 consumers saw no benefits from AI customer support[1][3][5]

2. **Oracle AI Agents Announcement (Oct 6, 2025)** - Oracle announced new role-based AI agents within Oracle Fusion Cloud Applications for marketing, sales, and service[4]

3. **Salesforce Agentforce Announcement (Oct 9, 2025 news about next week)** - Salesforce announcing two new capabilities: voice capabilities and hybrid reasoning for AI agents[7]

4. **Zendesk AI Agent (Oct 8, 2025)** - Zendesk launched new autonomous support agent claiming to solve 80% of support issues[13]

5. **NewDay AI Solution** - NewDay introduced AI-powered support solution saving 30 seconds per call[10]

Let me structure this as a news digest with 5th grade reading level, focusing on AI agents:

Key points to cover:
- AI agents are being widely deployed but facing challenges
- Major companies (Oracle, Salesforce, Zendesk) announcing new AI agent capabilities
- Consumer concerns about AI replacing humans
- The balance between AI and human support
- Real-world impact and results

I need to write in simple language (5th grade level), break into many paragraphs, use bold for key terms, and include citations.

Let me create the JSON structure with shortSummary (300-500 words) and longSummary (800-1200 words).
</think>

```json
{
  "shortSummary": "This weekly update brings important news about **AI agents** in customer service. Companies are racing to add **artificial intelligence** to their support teams, but customers are not always happy with the results.\n\nA major study from **Qualtrics** found that customer service quality is getting better overall, but **AI support is the weakest part**[1][3]. Nearly one out of every five people who used AI for customer help said it gave them **no benefits at all**[1]. Half of customers worry that AI will make it harder to talk to a real person when they need help[1].\n\nThe problem is that many companies are adding AI chatbots too quickly without making sure they work well[1]. Some companies fed old information from 2001 into their AI systems without cleaning it up first[1]. This led to **poor quality** conversations with customers[1].\n\nDespite these challenges, big technology companies are pushing forward with new AI agent tools. **Oracle** announced new **role-based AI agents** for marketing, sales, and customer service on October 6[4]. **Salesforce** revealed that its AI agents will soon be able to **speak like humans** and make smarter decisions[7]. **Zendesk** launched a new AI agent that claims to solve **80% of support problems** without human help[13].\n\nExperts say the best approach is using **AI and humans together**[2][5]. AI can handle simple questions like tracking orders or resetting passwords, while real people focus on **complicated problems** that need empathy and understanding[2]. Companies that try to replace humans completely with AI often damage customer trust[5].\n\nOne success story comes from **NewDay**, a financial services company that uses AI to help its customer service team[10]. Their AI system saves **30 seconds on each call** and helps the company handle 600,000 extra calls[10]. The key difference is that their AI **supports human workers** instead of replacing them[10].",
  "longSummary": "This weekly update covers major developments in **AI agents** and **agentic AI** for customer service. While artificial intelligence is changing how companies help their customers, the journey is not as smooth as many businesses hoped.\n\n**The Big Problem with AI Customer Support**\n\nA large study by **Qualtrics** released on October 7 painted a mixed picture[1][3][5]. The research looked at 20,000 consumers across 14 countries and found that overall customer experience quality is improving[1]. Customer satisfaction reached 79%, and trust levels hit 76%[1]. However, **AI-powered customer service** is struggling badly[1][5].\n\nNearly **one in five customers** who tried using AI for customer support said it provided **no benefits at all**[1][3]. Even more concerning, half of all customers worry that AI will block them from reaching a real human when they need help[1]. This shows a big gap between what companies want to achieve with AI and what customers actually experience[5].\n\n**Why AI Customer Service Is Failing**\n\nIsabelle Zdatny, a leader at **Qualtrics**, explained the main problem[1]. Many companies view customer support as something that costs money, so they rush to add AI chatbots to cut costs[1]. They think they can push customers away from expensive human agents and toward cheaper AI bots[1].\n\nThe trouble is these AI chatbots often **don't work very well**[1]. Companies deployed their AI systems **too quickly** without proper testing[1]. Some businesses fed their AI agents information from old databases without cleaning up the data first[1]. One example: AI systems were pulling policies from the year 2001 because they couldn't tell which rules were current and which were outdated[1].\n\nZdatny said companies \"got out a little bit over their skis,\" meaning they moved too fast[1]. This led to **really poor quality** conversations between AI bots and customers[1]. When customers have bad experiences with AI support, it costs companies money too. Half of bad experiences make customers **spend less** with those brands[1].\n\n**Major Companies Launch New AI Agent Tools**\n\nDespite the challenges, big technology companies announced exciting new AI agent products this week. On October 6, **Oracle** revealed new **role-based AI agents** built into their cloud applications[4]. These AI agents are designed to help with marketing, sales, and customer service tasks[4]. Chris Leone from Oracle said AI agents are **transforming customer interactions** from slow, manual processes into fast, proactive strategies[4].\n\n**Salesforce** announced two important updates coming to its AI agents[7]. First, the agents will gain the **power of speech**, allowing them to talk with customers over the phone and through websites[7]. Second, they will use something called \"hybrid reasoning\" that lets them be creative when needed but also follow strict rules for important decisions like giving refunds[7]. Adam Evans from Salesforce said companies need both flexibility and consistency[7]. The voice features will be available starting October 21[7].\n\n**Zendesk** joined the competition on October 8 with a new **autonomous support agent**[13]. The company claims this AI agent can **solve 80% of support issues** without any human help[13]. This is a bold promise in a crowded market where many companies are fighting for customers[7].\n\n**Success Stories and Best Practices**\n\n**NewDay**, a financial services company, shows how AI can work well when used correctly[10]. They handle **2.5 million calls** every year and added an AI-powered system to help their customer service team[10]. The AI saves an average of **30 seconds per call**, which equals the work of 10 full-time employees[10]. It also helps the company manage an extra 600,000 calls[10].\n\nThe key to NewDay's success is that their AI **supports human workers** instead of replacing them[10]. The AI gives instant help to customer service agents, speeds up training for new employees, and makes sure customers get faster service[10].\n\nExperts agree this is the right approach. Research shows that by 2025, AI will be involved in **95% of customer interactions**[2]. Companies using AI well report **17% higher customer satisfaction**, **30% lower costs**, and even **4% more revenue growth**[2]. But the winning strategy is **AI plus human agents working together**[2][5]. Machines handle repetitive tasks while people focus on empathy, nuance, and strategy[2].\n\nMatt Trickett from Qualtrics said it simply: \"Companies that design AI to support the relational, not just enhance the transactional, are offering the best customer experience\"[5]. This means AI should help build relationships with customers, not just complete transactions quickly[5].\n\n**Looking Ahead**\n\nThe future of AI in customer service depends on finding the right balance[2][5]. Customers want fast answers, but they also want to feel understood[2][5]. They expect personalized help, but they worry about privacy when AI systems use their data[5]. Companies that listen to these concerns and use AI to **support humans** rather than replace them will likely succeed[2][5]. Those that rush to cut costs with poorly designed AI chatbots risk losing customer trust and money[1][5]."
}
```

CX quality is improving, no thanks to AI customer support

The Future of AI in Customer Support: Trends for 2025 & ...

AI-Powered Customer Service Fails at Four Times the Rate ...

Oracle AI Agents Help Marketing, Sales, and Service ...

Global Consumer Experience Trends for 2026 Revealed

AI-powered next best experience for customer retention

Exclusive: Salesforce lets AI agents speak for themselves

10 Top CRM Trends for 2025 to Consider

The Bright Spot for AI ROI in 2025 Is Customer Service

AI in the portfolio: NewDay customer service solution - CVC

Global Consumer Predictions: 2026 & Beyond

12 Proven Examples of AI in Customer Service in 2025

Zendesk says its new AI agent can solve 80% of support ...

People and Planet Consumer Insights and Trends 2025 ...

AI Customer Research: Practical 2025 EU Guide for CPG

Be My Eyes Announce Launch of Service AI as a ...

WARC Consumer Trends 2025: Five shifts shaping ...

Forge 2025: Advancing customer engagement in an AI- ...

CX leaders need to drive AI governance before it's too late

The Gen Z paradox: Spending less, expecting more

AI Adoption to Customer Loyalty - Retail Trends for 2025

[FULL] Consumer trends that will shape the world

Customer care and network support are driving telco AI ...

Future of Commerce: Americas Edition

This weekly update covers major developments in **AI agents** and **agentic AI** for customer service. While artificial intelligence is changing how companies help their customers, the journey is not as smooth as many businesses hoped.

**The Big Problem with AI Customer Support**

A large study by **Qualtrics** released on October 7 painted a mixed picture[1][3][5]. The research looked at 20,000 consumers across 14 countries and found that overall customer experience quality is improving[1]. Customer satisfaction reached 79%, and trust levels hit 76%[1]. However, **AI-powered customer service** is struggling badly[1][5].

Nearly **one in five customers** who tried using AI for customer support said it provided **no benefits at all**[1][3]. Even more concerning, half of all customers worry that AI will block them from reaching a real human when they need help[1]. This shows a big gap between what companies want to achieve with AI and what customers actually experience[5].

**Why AI Customer Service Is Failing**

Isabelle Zdatny, a leader at **Qualtrics**, explained the main problem[1]. Many companies view customer support as something that costs money, so they rush to add AI chatbots to cut costs[1]. They think they can push customers away from expensive human agents and toward cheaper AI bots[1].

The trouble is these AI chatbots often **don't work very well**[1]. Companies deployed their AI systems **too quickly** without proper testing[1]. Some businesses fed their AI agents information from old databases without cleaning up the data first[1]. One example: AI systems were pulling policies from the year 2001 because they couldn't tell which rules were current and which were outdated[1].

Zdatny said companies "got out a little bit over their skis," meaning they moved too fast[1]. This led to **really poor quality** conversations between AI bots and customers[1]. When customers have bad experiences with AI support, it costs companies money too. Half of bad experiences make customers **spend less** with those brands[1].

**Major Companies Launch New AI Agent Tools**

Despite the challenges, big technology companies announced exciting new AI agent products this week. On October 6, **Oracle** revealed new **role-based AI agents** built into their cloud applications[4]. These AI agents are designed to help with marketing, sales, and customer service tasks[4]. Chris Leone from Oracle said AI agents are **transforming customer interactions** from slow, manual processes into fast, proactive strategies[4].

**Salesforce** announced two important updates coming to its AI agents[7]. First, the agents will gain the **power of speech**, allowing them to talk with customers over the phone and through websites[7]. Second, they will use something called "hybrid reasoning" that lets them be creative when needed but also follow strict rules for important decisions like giving refunds[7]. Adam Evans from Salesforce said companies need both flexibility and consistency[7]. The voice features will be available starting October 21[7].

**Zendesk** joined the competition on October 8 with a new **autonomous support agent**[13]. The company claims this AI agent can **solve 80% of support issues** without any human help[13]. This is a bold promise in a crowded market where many companies are fighting for customers[7].

**Success Stories and Best Practices**

**NewDay**, a financial services company, shows how AI can work well when used correctly[10]. They handle **2.5 million calls** every year and added an AI-powered system to help their customer service team[10]. The AI saves an average of **30 seconds per call**, which equals the work of 10 full-time employees[10]. It also helps the company manage an extra 600,000 calls[10].

The key to NewDay's success is that their AI **supports human workers** instead of replacing them[10]. The AI gives instant help to customer service agents, speeds up training for new employees, and makes sure customers get faster service[10].

Experts agree this is the right approach. Research shows that by 2025, AI will be involved in **95% of customer interactions**[2]. Companies using AI well report **17% higher customer satisfaction**, **30% lower costs**, and even **4% more revenue growth**[2]. But the winning strategy is **AI plus human agents working together**[2][5]. Machines handle repetitive tasks while people focus on empathy, nuance, and strategy[2].

Matt Trickett from Qualtrics said it simply: "Companies that design AI to support the relational, not just enhance the transactional, are offering the best customer experience"[5]. This means AI should help build relationships with customers, not just complete transactions quickly[5].

**Looking Ahead**

The future of AI in customer service depends on finding the right balance[2][5]. Customers want fast answers, but they also want to feel understood[2][5]. They expect personalized help, but they worry about privacy when AI systems use their data[5]. Companies that listen to these concerns and use AI to **support humans** rather than replace them will likely succeed[2][5]. Those that rush to cut costs with poorly designed AI chatbots risk losing customer trust and money[1][5].

This weekly update brings important news about **AI agents** in customer service. Companies are racing to add **artificial intelligence** to their support teams, but customers are not always happy with the results.

A major study from **Qualtrics** found that customer service quality is getting better overall, but **AI support is the weakest part**[1][3]. Nearly one out of every five people who used AI for customer help said it gave them **no benefits at all**[1]. Half of customers worry that AI will make it harder to talk to a real person when they need help[1].

The problem is that many companies are adding AI chatbots too quickly without making sure they work well[1]. Some companies fed old information from 2001 into their AI systems without cleaning it up first[1]. This led to **poor quality** conversations with customers[1].

Despite these challenges, big technology companies are pushing forward with new AI agent tools. **Oracle** announced new **role-based AI agents** for marketing, sales, and customer service on October 6[4]. **Salesforce** revealed that its AI agents will soon be able to **speak like humans** and make smarter decisions[7]. **Zendesk** launched a new AI agent that claims to solve **80% of support problems** without human help[13].

Experts say the best approach is using **AI and humans together**[2][5]. AI can handle simple questions like tracking orders or resetting passwords, while real people focus on **complicated problems** that need empathy and understanding[2]. Companies that try to replace humans completely with AI often damage customer trust[5].

One success story comes from **NewDay**, a financial services company that uses AI to help its customer service team[10]. Their AI system saves **30 seconds on each call** and helps the company handle 600,000 extra calls[10]. The key difference is that their AI **supports human workers** instead of replacing them[10].

Customer Service Weekly AI News

Specific Topics