In this episode, Chris Hague, SVP at Bill Gosling Outsourcing, discusses the role of AI in the BPO and collections industry.
The conversation explores the evolution of AI from RPA-style automation to large language models, cultural and operational challenges in adoption, client risk concerns, and the balance between efficiency and customer experience.
Key themes include the shift from cost-per-seat to cost-per-outcome, the future of ring-fenced models, the importance of risk management, and the broader societal impact of automation.
The discussion positions AI as both a transformative enabler and a frontier filled with challenges requiring structured, controlled approaches.
Find out more about Bill Gsoling -> Here.
Key Take Aways
- AI development is advancing rapidly, with large improvements seen even within six months.
- There remains confusion between AI, RPA, and workflow automation, with many implementations still closer to programmed processes than autonomous AI.
- Organisations need cultural readiness and dedicated transformation teams to adopt AI effectively.
- Some of the best automation ideas originate from frontline staff once they understand the direction of change.
- Risk management and customer data protection are critical client concerns in AI deployment.
- Generational differences shape acceptance, with younger users more comfortable using AI self-service tools.
- AI’s role is strongest in transactional and repetitive tasks, but human empathy remains essential for complex and vulnerable cases.
- Ethical questions persist around transparency—whether customers should always know if they are interacting with AI.
- The industry is experimenting with private/local models to address consistency, risk, and client-specific requirements.
- Prompt testing may emerge as a formal discipline, similar to penetration testing in cybersecurity.
- BPO models may shift from cost-per-seat to cost-per-outcome as AI changes operational economics.
- The BPO industry must continue evolving, balancing efficiency with customer experience and risk controls.
Innovation
- Use of narrow pilots and test-and-learn approaches before scaling AI adoption.
- Exploration of multi-model “chunking” to reduce latency in chatbots.
- Potential ring-fenced client-specific AI models to protect data and ensure consistency.
- Development of prompt testing as a safeguard against misuse or prompt injection.
- “Super agent” functionality, combining knowledge across multiple areas into one seamless customer interaction.
Key Statistics
- Recruitment processes automated up to 18 months ago, narrowing thousands of CVs into manageable pools with human oversight.
- Frontier for AGI seen as 5–10 years away, though progress is accelerating.
Key Discussion Points
- Definition and perception of AI vary widely, shaping adoption strategies.
- Transition from process automation to intelligent digital operations.
- Balancing operational efficiency with customer experience.
- Importance of cultural buy-in across all organisational layers.
- Evolution of customer trust and acceptance of AI.
- Risk of reputational damage if AI outputs are untrustworthy.
- Differentiation of transactional vs. empathetic use cases in customer contact.
- Ethical considerations on transparency of AI vs. human interaction.
- BPO industry transformation under AI pressures.
- Client risk appetites dictate AI deployment speed and model choices.
- Societal implications of automation on labour markets and job design.
- Frontier mindset—progress will be messy but necessary to capture value.
#BillGosling
RO-AR insider newsletter
Receive notifications of new RO-AR content notifications: Also subscribe here - unsubscribe anytime
