AI is making Philippine call center work more efficient, for better and worse

From the article:

In the Philippines, advanced AI tools — including those with language recognition, emotion recognition, and generative intelligence — have made work more demanding, six BPO workers and a representative from a BPO worker’s association told Rest of World. They spoke of targets rising since 2022 and fears of industry layoffs and redundancies.

The Philippines, the second-largest BPO market in the world after India, has 1.84 million BPO workers. Although there is no official data for job losses due to AI, the Philippines’ labor secretary, Bienvenido Laguesma, told local media in June that some workers are already losing their jobs to AI. Industry estimates suggest that while 300,000 Filipinos could be out of work due to AI in the next five years, 100,000 new jobs could be created in roles like data curation.

The Philippines leads the world in AI adoption, and 86% of Filipino white-collar workers already use AI to “boost productivity, efficiency and creativity,” according to the 2024 Work Trend Index created by LinkedIn and Microsoft. Two-thirds of BPO companies that are members of the IT & Business Process Association of the Philippines are already using AI or are piloting it, Dominic Ligot, the association’s head of AI and research told Rest of World.

Even companies that prioritize a human touch are forced to use the tech to satisfy clients who are demanding greater automation, Alex Peña, director of special projects at Boldr, a BPO firm recognized by the nonprofit B Lab Global for its social impact, told Rest of World.

...

“Ideally, AI would be helpful. But what’s happening is that companies are using it to justify adding more tasks,” Porquia said. “At once, one can do customer service, sales, and tech support.”

Porquia said Filipino BPO workers are at the lower end of the tech value chain, performing simple, repetitive tasks that can be automated.

...

The BPO employees Rest of World spoke with said AI co-pilots made them more efficient. They said the program recognizes what is said, swiftly pulls up the customer’s past concerns, and suggests solutions and follow-up questions in real time. The Concentrix employee who requested anonymity said the co-pilot even “tells me if I need to slow down, speed up, or deliver a statement with empathy.”

...

The employee is now a trust and safety analyst at Accenture, where she checks the accuracy of AI-generated data for Facebook’s parent company, Meta. She said she is working on an upcoming feature wherein Meta AI recognizes photos posted to Facebook and Instagram and displays relevant prompts next to them for the user to explore. She said she cross-checks the Meta AI’s response for accuracy, with the help of a Microsoft AI assistant. Another employee, who requested anonymity due to a nondisclosure agreement, said she works on a similar program for Instagram reels. A Meta spokesperson declined to comment.

The employee working on Facebook said that two months ago she was given 200 seconds per prompt.

This was revised to 170 seconds per prompt in October.

“[The managers] check our errors and say we can’t drop below a 90% accuracy,” she said. “It’s getting really tough; hardly anyone can keep up consistently. You see others in your team scoring just 60% and worry for them.”