Consultancy firms and banks warned that generative AI, which can process human language, could be upgraded to the point where humans are automated out of work. One Goldman Sachs report in March 2023 said 300 million jobs could be exposed to automation. About 46% of those are office and administrative jobs like customer service.

Millions of these jobs are found in developing economies in Southeast Asia like that of the Philippines, which means AI development could end up derailing economic growth.

But more than two years after ChatGPT made its debut, generative AI developments have not yet produced the feared doomsday scenario. There have been job losses, but the industry still depends on its human workers, and at this point AI is just another tool companies use to augment their capabilities.

Renz Miguel Marquez, a 29-year-old call center trainer in Makati City in the Philippines, uses AI tools to streamline his workflow. The call center where he works caters to broadband service transactions for a U.S.-based company.

Marquez says his call center introduced an “agent assist” about two years ago that listens to the conversation and helps agents collate information that could be useful to help resolve issues. That tool has helped speed up resolutions, giving agents more time to pitch and sell services or products to customers who call in.

Marianne Capitly, a 32-year-old in Manila, handles after-sales support for a U.S.-based IT distributor that sells computer parts, laptops, and software. Most of her work is email-based, and she said new AI applications have streamlined her work as she searches for solutions either on the internet or in an internal database.

Nominations are now open:
Fortune is now accepting nominations for the 2025 Southeast Asia 500—the definitive ranking of the region’s largest companies. Start your nomination here.

Both agree the technology makes them more efficient, but it isn’t ready to replace them. The tools are still “not quite there yet,” they both say, and can churn out solutions that are not applicable.

While the future impact on service jobs from AI is still very much unknown, it may not be as dire as first predicted. A recent report from Oxford Economics, an economic advisory firm headquartered in the U.K., argues productivity gains from generative AI developments will skew toward services and boost medium-term economic growth. The business process outsourcing (BPO) industry is one sector that could benefit, allowing workers to transition into other customer-facing roles. That could be comforting news for some 1.8 million people employed by the approximately 1,000 BPO companies in the Philippines, which together contribute about 8% of the country’s annual GDP.

One key development arguably helped grow the BPO sector in the early 2000s: cheaper long-distance phone calls.

“The real revolution was around telecommunications,” says Laurent Junique, CEO of TDCX, a Fortune Southeast Asia 500 company and provider of BPO services. The “democratization” of the telecom sector lowered the cost of calls from the U.S. to the Philippines or India, making offshoring feasible. It was also in the early 2000s that major firms like Accenture set up BPO centers in the Philippines.

The Philippines’ BPO sector is expected to have generated around $38 billion in revenue last year, according to the IT and Business Process Association of the Philippines (IBPAP), a trade group which represents about 400 companies.

While BPO services include tasks like payroll processing, accounting, and even social media management, for the past two decades or so, the BPO sector in the Philippines has been associated with call centers serving English-speaking customers.

For Junique, AI developments are just the latest change in the industry. AI tools enable agents to perform higher-level work such as conflict resolution for unhappy customers, he says. Such work could potentially allow TDCX to charge clients more as it provides a more complex suite of services.

AI is part of a continual evolution in the sophistication of services BPO providers can offer clients. For example, Junique says 10% to 15% of the calls TDCX used to get for one airline were for account password resets, but that process had been automated a long time ago.

“There’s been several waves of automation and simplification of processes and leveraging all the advanced tools possible to navigate toward getting the more complex work and retiring the more simple tasks,” Junique says.

As AI-driven tools become more prevalent, BPO companies are likely to develop and offer new services, providing additional revenue streams, according to the International Monetary Fund (IMF).

The results of a survey conducted by the IBPAP released in November show incorporating AI tools may already be underway, with 56% of companies actively integrating AI.

But to be sure, technological advancements will still result in job losses.

The same IBPAP survey estimates about 8% of companies have cut headcount, likely owing to the automation of specific tasks.

An IMF spokesperson said 14% of the total Philippines workforce could also be at risk of being replaced by AI, and investments in education and digital infrastructure are needed to leverage the future of work. To guard against job losses, the Philippines government has developed a National AI Strategy Roadmap, which aims to integrate AI across various sectors and prepare the workforce for shifts in the job market.

Junique concedes there will be job losses. But he also predicts that as AI becomes more commonplace, the BPO sector will expand to provide tech support in new sectors. Self-driving cars, he says, will need agents to handle queries from customers, and maybe even to sell them.

“Before, you’d buy your cars from a dealer; now [dealers] are going to come sit in our centers because cars are bought online,” Junique says.

This article appears in the February/March 2025 Asia issue of Fortune with the headline: “AI may not take our jobs after all.”