'CIOs are under increasing pressure to generate business value from generative AI, but Salesforce CIO Juan Perez is among those IT leaders who believe such impatience from the C-suite could doom many projects.

“The explosion of AI has really put CIOs in the spotlight and placed the CIOs in the hot seat like never before,” says Perez, fresh off two industry events in which many of his CIO colleagues say they feel they are between a rock and a hard place. “It is really clear to me how CIOs and other tech executives from all kinds of organizations are being challenged as to how they can to move quicker, but also responsibly, with AI.”

The enterprise landscape is littered with Version 1.0 generative AI proof of concept projects that did not materialize into business value and have been dumped. While the industry remains in the early stages of uncovering and implementing AI use cases into business workflows, the blueprint for success remains elusive for many CIOs.'

NB: when insurers talk about GenAI I feel that they are mixing up all kinds of AI and lumping them together but I'll leave you to consider that.

A recent study by EY Parthenon indicates the high level of Generative AI expectations across the insurance industry (link to full report in Further Reading below). 

This supports Perez's view that the C-Suite is pressuring CIOs to experiment and much of this in customer-facing service, chatbot use cases. Future priorities such as claims and underwriting are expected to deliver benefits over the next decade. 

Source EY Parthenon: Generative AI in Insurance May 2024

 There are four hurdles to overcome to transition from experiments to a long-term process of constant innovation, improvement, and achieving the goals the C-Suite longs for.

  1. Replacing legacy core systems with modern cloud-native, AI-native, API-rich, MACH-architected (see appendix) cores that offer data fluidity and enable connected ecosystems
  2. Put data management and maturity back in vogue ( see Further Reading for full discussion). 
  3. Answer the WHY? Ensure there is a clear corporate vision backed by the strategies, resourcing, and cross-company buy-in that leverages AI to achieve the enterprise goals (See ‘Prioritising Output Goals’ in Further Reading).
  4. Finding the right technology partners that understand and will operationalise data and AI collaboratively with internal teams ( See Further Reading- ‘What makes an optimal technology partner’)

From the CIO perspective, Perez offers answers.

"But much of the business value the C-suite is clamoring for cannot be rushed. For CIOs, the challenges ahead are staggering, but creating a comprehensive enterprise architecture with AI assessments, as well as security and data assessments, will be critical to innovation.  

“I believe AI will have a positive impact in lowering and managing costs, which in turn helps organizations create value,” Perez says. “There are a lot of misconceptions as to what is the appropriate use of generative AI and misunderstanding as to the areas in which you can create the most value from AI.”

Follow the links below to read the full article by Perez, the EY Parthenon report on GenAI, tackling data, and setting priority outcome goals for GenAI. 

Further Reading

CIOs under pressure to deliver AI outcomes faster by Paula Rooney in CIO

Generative AI in insurance report by EY Parthenon

Surveys highlight data maturity holding back AI deployment ambitions

Prioritising outcome goals leveraging GenAI- productivity/cost-cutting, increasing revenue, transformative competitive advantage, or all of these?

What makes an optimal technology partner?

Appendix  MACH architected cores

What is a MACH-architected system?

  • Microservices 
  • API first
  • Cloud native
  • Headless

Microservices architected means that you can build a software product as a set of independent components — microservices — where each component operates on its own and interacts with others through APIs. As a result, teams can deploy, change, and improve separate software components without disrupting the rest of the system.

API-first architectures are more flexible allowing teams to choose the most appropriate frontend technology to solve priority business problems. Developers can unify logic across touchpoints and avoid duplication of development work, as well as eliminate channel silos.

APIs allow for fast communication between components, meaning that businesses can reduce the time to implement new touchpoints and accelerate speed-to-market processes.

Cloud-native platforms use the public, private, and hybrid cloud as part of a cloud migration strategy to develop scalable and dynamic data solutions. Insurers will be more resilient to performance issues that often bug on-premises systems.

Headless - far from being clueless!  Insurers that employ headless architecture don’t have a default frontend system that defines how content is presented to end users. You will be able to deliver personalised products and services to your target audience using any channels, devices, and platforms.