Microsoft's Copilot Studio and Outsourcing Economies

As artificial intelligence continues to evolve, tools like Microsoft’s Copilot Studio are transforming how businesses manage their workflows, particularly in areas that have traditionally relied on human labor. This transformation poses a significant threat to countries heavily dependent on outsourcing jobs, such as customer service, content moderation, and sales. These roles have been a primary source of income for many economies around the world, including India, the Philippines, Vietnam, Bangladesh, and Egypt.

Advancements in AI technologies are addressing previous concerns about the reliability and effectiveness of LLMs in handling these tasks. With features like triggered and templated responses, training on workflow-relevant documents, and escalation to human agents when needed, AI tools are becoming increasingly capable of performing outsourced jobs, at least to the extent of decreasing the workload significantly for human agents, if not yet able to replace them fully. making it critical to understand the implications for countries that rely heavily on outsourcing.

Microsoft Offices.

Automating Human-Driven Workflows

Microsoft’s Copilot Studio is designed to automate workflows that involve repetitive human interactions. This includes tasks typically performed by outsourced professionals, such as:

• Customer Service: AI-driven chatbots and virtual assistants handle customer inquiries and resolve issues without human intervention, equipped with training on documents relevant to the business operation as "Knowledge Sources" in Microsoft terminology. Also enabling programmers to engineer answers on certain "Triggered Topics", as to add a layer of control over the copilot's interaction with customers.
• Content Moderation: Automated tools review and flag user-generated content for compliance with community guidelines, diminishing the demand for human moderators. Which is one of the strong suits of LLMs generally, now the guidelines can be added as "Knowledge Sources" to the model, and it will handle the rest.
• Sales Roles: AI assists in lead generation, customer follow-ups, and other routine sales tasks, through the use of automated actions that can be engineered to get triggered at certain points in the interaction. thereby threatening outsourced sales teams that handle repetitive, script-based interactions.

A Call Center in Egypt.

Economic Dependence on Outsourcing

Certain countries are particularly vulnerable to the effects of this shift toward automation due to their heavy reliance on outsourcing:

1. The Philippines: The Philippines has now become the largest offshore voice-related call center market with more than 400,000 call center workers, according to the Contact Center Association of the Philippines. The overall industry, including the other back office-related functions, has expanded to more than 1 million workers. Just last year, companies leased more than 5 million square feet of office space for back office and call centers including one of the biggest deals of the year by Citibank, which leased more than 500,000 square feet for an estimated 7,000 workers. (Site Selection Group Blog).
2. India: While India is known for its IT outsourcing, a significant portion of its $194 billion revenue from the IT-BPO industry in 2022 comes from non-IT services, including customer support and call centers. There were approximately 350,000 call center workers in India as of February 2015. (Site Selection Group Blog).
3. Vietnam: The country's growing BPO sector includes substantial customer service and sales support operations, contributing to its overall $9 billion outsourcing revenue in 2021 ( Vietnam Software Association, 2021 Report).
4. Bangladesh: With around 300,000 freelancers, many in roles like customer service and sales, Bangladesh’s outsourcing industry is projected to reach $5 billion by 2025 (Bangladesh Association of Software and Information Services, 2021 Report).
5. Egypt: Egypt’s outsourcing sector, which focuses heavily on call centers and customer support roles, generated $4.2 billion in 2020, employing around 174,000 professionals ( Egypt’s Information Technology Industry Development Agency, 2020 Report).
6. Jordan: The outsourcing sector in Jordan has been growing steadily, with the country's Business Process Outsourcing (BPO) and Information Technology Outsourcing (ITO) sectors contribute around $200 million annually to the economy. This sector employs approximately 5,000 people, primarily in customer service and content moderation roles (Jordan’s Information and Communications Technology Association, 2023 Report).

However, as of the writing of this article, Copilot is yet to support Arabic. So, Egypt and Jordan are safe for now.
Not to mention the challenge to support all Arabic dialects, not just Fus-ha Arabic, since not all Arabs speak Fus-ha.

Baguio, Benguet, Philippines

Economic and Social Implications

The automation capabilities of tools like Copilot Studio pose several challenges:

• Job Losses in Key Sectors: As companies adopt AI-driven automation tools, the demand for outsourced automatable roles is likely to decline, potentially resulting in substantial job losses in countries like India and the Philippines.
• Economic Disruption: A reduction in outsourcing demand could disrupt the economies of these countries, leading to slower economic growth, increased unemployment, and potential socio-economic crises. And with barely any social safety net for people to fall back on, their humanitarian situation could potentially worsen.
• Increased Inequality: As AI-driven automation reduces the need for human labor, profits for companies are likely to increase. In countries with non-progressive tax systems, or commonplace tax evasion, these increased profits could exacerbate income inequality, concentrating wealth among those who own or control the technology.

Weighing Potential Mitigation Strategies

To counter these potential adverse effects, countries heavily reliant on outsourcing should consider the following strategies:

• Diversification of the Economy: Investing in other sectors such as manufacturing, tourism, and local IT development to reduce dependency on outsourcing.
  However, in most cases, such countries find themselves in a situation where they rely on outsourcing because in fact they have significant social problems preventing the development of such industries, for example, corruption, or dictatorship.
• Upskilling the Workforce: Training workers in areas less likely to be automated, such as advanced AI development, creative industries, and high-level software engineering.
  On the other hand, training at such a high level requires the development of the education sector, supporting relevant start-ups, and funding research in relevant fields.
  As to incentivize the creation of a tech and creativity echo system. Realistically, this is all very unlikely to happen in a developing country.
• Investing in Domestic Technology Development: Encouraging local technology startups and innovation hubs to create a more resilient and diversified economy.

LLMs are not the super intelligent AIs we see in movies. But efforts to capitalize on their strengths, are already making us face the type of changes, we usually associate with the arrival of Artificial General Intelligence (AGI). Most importantly, socio-economic changes in vulnerable countries.