From Vision to Reality: Four Pillars for a Robust AI Strategy

AI’s power can be transformative—a well-rounded plan for deploying it can help.

Even though it’s still early days in the age of artificial intelligence (AI), it sometimes feels like this innovation has been around forever. In that short time, the pace of change has been staggering. It may seem daunting to lay out a long-term AI strategy for an organization, but it’s critical.

The Starting Point: What’s Your Risk Appetite?

We think a good starting point is to pin down how far your organization is willing to go with such transformative technology.

Each firm may have a different trade-off for opportunity versus risk that informs the strategy. Making that determination requires asking fundamental questions. What is your organization comfortable with? What types of risks are acceptable? Those answers set the tone for how AI tools are used. Scope is another key element: Should AI tools be used throughout the organization or only in certain functions?

Organizations must also set the standard for the roles humans and machines play in carrying out key functions, guided by a philosophy on how AI and humans coexist. From our perspective, pairing human expertise with machine intelligence is an effective way to create augmented intelligence, with Iron Man as the inspiration. In other words, we integrate a tech-savvy human expert with an AI-powered “suit of armor” that delivers timely, effective prompts to enable decisions.

Defining an Organizational Strategy for AI

As we developed our AI strategy, we found it useful to organize it around four pillars that address which problems we tackle with AI, how we build out technological and human capabilities, and how we make sure we use AI in a responsible way.

1. Surface and prioritize opportunities. AI should be focused on areas where it can really have impact—use cases that matter. Those applications could be in any corner of the organization: A tool for improving investment performance. A more precise way to manage risk. A more streamlined operational process. More effective experiences for clients.

Creating a stock primer is a strong AI use case because it directly addresses a core analyst problem: quickly getting to a relevant, decision-ready context in a world of information overload. By automatically synthesizing fundamentals, recent news, risks and debates, AI surfaces and prioritizes opportunities that might otherwise be missed—or require hours of manual work to uncover.

2. Build foundational infrastructure and capabilities. With AI, a strong foundation is a prerequisite for broader adoption. A robust and diverse set of shared tools, data connectors and resources enables the organization to leverage AI at scale.

For example, we’ve built a platform that enables employees to use AI safely, empowering the firm to leverage the latest tools while protecting client data and our intellectual property. Generative AI tools extend a foundational capability across departments with very different functions. AI applications need to talk to other systems, so efficient connections are critical. Application programming interfaces (APIs) allow developers to create, integrate and manage AI infrastructure.

3. Develop AI talent and dexterity. Employees need to develop a symbiotic relationship with AI. Many people are familiar with AI tools like ChatGPT. They may even be power users in their personal lives, researching their favorite movies, understanding what they need for their tax returns or shopping for a car. But tailored training is essential for AI use cases that support different professional roles in an organization, whether it’s investing, client service or operations.

Hands-on training is invaluable—especially from peers. It gives employees a window into the tangible benefits of using AI in their daily work and tasks. Sharing examples of prompt engineering. Friendly contests on AI use cases. Live demos. We’re using many approaches to tailor AI dexterity. And we’re encouraging colleagues to experiment. Individuals are very imaginative in exploring ways that AI might help them, and they have the domain expertise and boots-on-the-ground experience to judge whether it’s working.

4. Do it responsibly. AI’s power holds enormous promise, but that power also brings risks. Organizations must use AI ethically and transparently, with strong controls and accountability measures. A strong governance framework can help achieve these objectives and serve as an in-house supervisory function. Its remit includes vetting all AI use cases and monitoring them to ensure that tools are performing as expected and employees are using them as they’re intended.

Guidance and controls should be designed to ensure that AI tools complement human judgment and experience, not replace them, a need amplified with the onset of agentic AI. AI may be able to help identify potential portfolio risks, but humans must make the final decision on whether to act, because they can consider factors and context that AI might not fully grasp. Human-centricity promotes the effective and responsible use of AI tools. And because data are an essential ingredient for AI, tools must be grounded in reliable and secure data sources.

From a big-picture perspective, the age of AI is still in its very early stages, but it’s rapidly transforming into a business imperative. That’s why it's critical for organizations to develop and implement an AI strategy as part of their overall business strategy.