How soon before the Call Center Industry is completely disrupted by automated AI Agents?

AI is everywhere.  Humanoids are coming.  And soon enough your local call center employees will not be needed.  Hyperbole?  Perhaps, but consider the following references and follow along with my thinking:

Pictory is a simple service that creates video from text prompts.  It may not be in the same league as Adobe Firefly, OpenAI’s Sora and competing offerings by Meta and others, but it works and at this moment it illustrates the capabilities of AI voice talent.  Pictory offers over 100 unique voice styles, male and female, selected by name, language, accent, age, purpose (e.g. childrens’s stories, meditation) and gender.  Users can select Standard voices or high definition Premium voices.  And they sound reasonably realistic.

For inbound order taking, who needs to hire a thousand inbound agents, when you could possibly scale up digitally with autonomous inbound AI agents.  It’s all software, AI and large data centers.  No hiring or firing required.

In the United States, Statista reports 2.86 million contact center employees.  I’ll estimate that 50% or more of these are live people on the phone.  This is based on my ten years or so in the outsourcing industry and consulting with call center companies.  Within inbound contact centers the ratio may be higher but let’s go with 50%.  1.4 million people on the phones in the United States talking to customers in identified contact centers.  U.S. contact center industry: employees 2023 | Statista

And how many more in the Philippines? Call Centers Philippines: Revolutionizing customer service through advanced tech (businessnewsasia.com)  At least 1.3 million call enter employees by some estimates in 2019 with annual growth of 8-10%, but we’ll stay conservative and use the 1.3 million number. Philippine Call Center Industry Trends in 2023 and Beyond (superstaff.com)

Furthermore, if we consider individual companies, here are a few prominent players: 

Concentrix, founded in 2006.  One of the established leaders in BPO services offers marketing, IT, and customer service. They merged with Convergys in 2019, boosting its worldwide talent base to 265,000+ worldwide, 80,000 of them based in the Philippines.

Sutherland, a another large BPO / contact center company with 40k+ employees globally helps bring home the point of where I am going with my comments:

“Artificial Intelligence. Automation. Cloud engineering. Advanced analytics.
For business leaders, these are key factors of success. For us, they’re our core expertise.”

And let’s add one more simply for numerical reference, TTEC:  58,000 employees and counting on six Continents.  https://www.ttec.com/sites/default/files/2024-07/ttec-2022-q1-fact-sheet.pdf

Using the above information GenAI large language model training can occur in large scale across tens of thousands of call center agents, and across millions of calls annually.  GenAI providers are capable of creating custom models for individual companies.  Imagine this as a fairly straight forward project within any one of these large BPO companies:  Record all phone conversation within each call center, train the GenAI LLM on the recorded conversations, select the desired voice talent (language, accent, purpose, etc), deploy the model with human-in-the-middle processes and then migrate to full AI autonomy for inbound order taking and customer service.

McDonald’s recently completing testing the use of AI for order taking in their drive through restaurants.  I’m guessing, like call centers, they have the ability to record millions of conversations in which to create the content to create custom models. Although the recent test showed mixed results they are continuing to search for ways to implement AI on a broad basis.

Banks, airlines and others already use automated attendants to provide everything from flight status, to “where’s my bag” updates to “what’s my current balance” responses.  “One moment please and I will have that for you…” These are simple examples, but early indications of what AI is capable of automating.

Within our own company we offer custom, GenAI models targeted to traditional print publishers, using material contained in their archives; combined with LeadSticker, our two-way conversational SMS platform, while adhering to human-in-the-middle practices.  Think Scientific American and Journal of the American Chemical Society.  We also deploy full telephone record capability, although clearly at a much smaller scale than what has been referenced for the large BPO’s above.

Here’s my point: our company is small by comparison and yet the AI tools available today make enormous capabilities possible.  Click Here or on the image below to see video of an AI created Call Center Agent handling an inbound call to buy a new television.  It is simple and rudimentary, but consider what can be done by large companies with large resources:

“Artificial Intelligence. Automation. Cloud engineering. Advanced analytics.”

 In 2023, the United States opened fewer call centers in 2023 as the number of contact center employees decreased between 2020 and 2023. Number of Call centers opened or expanded by region 2023 | Statista.  “In 2023, there were roughly 2.86 million people working in contact centers, a decrease when compared to the previous year.” U.S. contact center industry: employees 2023 | Statista

Back to my headline question: How soon before the Call Center Industry is completely disrupted by automated AI Agents?

I do not have an answer to the question, and there are others within the call center industry with far more insight than what I am offering.  However, the technology and its capabilities are not lost on me in the realm of customer interactions and how soon it will be very difficult to know when you are talking with a live person and when it just might be artificial intelligence.