Organizations Master Technology. But AI Transforms Organizations.

Organizations have mastered the implementation of technology. AI requires them to master the transformation of themselves.

Why is that? Why is AI’s nature such that it is stretching the typical assessment process of IT initiatives beyond what this process was designed to do?  Why are so many organizations perplexed by the new AI paradigm?

We are still asking yesterday's questions

For decades, organizations have mostly followed a familiar pattern when evaluating IT demands successfully. These demands generally went through an assessment process centered on two principal questions:

·        Does the initiative make business sense?

·        Is the technology solution feasible, secure and affordable?

This approach has served organizations well because traditional IT primarily automated existing business processes without fundamentally changing how organizations made decisions or exercised accountability.

AI is fundamentally different. Unlike traditional IT, AI increasingly participates in (and takes over) activities that were previously the exclusive domain of people:

• analyzing information.
• generating recommendations.
• producing content.
• influencing decisions.
• interacting with customers.
• assisting professionals.
• and, in some cases, making autonomous decisions.

The central question "can we implement this technology?” typically boiled down to answering the more direct questions as follows:

• Is there a business case?
• Can we build it?
• Can we integrate it?
• Can we secure it?

While these questions are still pertinent and necessary, they are simply no longer sufficient to answer the fundamental new question reflecting the reality of introducing AI: "How will this technology change the way our organization operates?"

AI introduces entirely new organizational questions concerning:

• leadership.
• accountability
• governance
• ethics
• trust
• organizational culture
• workforce capabilities
• decision authority.

AI doesn't simply automate work, it changes organizations.

AI increasingly influences how decisions are made, how accountability is exercised and how leadership is practiced.  This overhaul of traditional dynamics within organizations demands an in-depth review of issues taking into account AI’s newfound potential status as a substitute to well-established organizational processes.

A simple example

Let us use a concrete example of the breadth of effects that an otherwise seemingly straightforward AI initiative can have on an organization.

Consider one of the simplest AI initiatives an organization can undertake: implementing AI-based automated meeting minutes-taking.  This already is a widespread practice in many organizations.  When analyzing the potential impact of such an initiative, it quickly becomes apparent that it extends far beyond the technology itself. The table below illustrates how many organizational areas/departments are potentially affected.

This pattern of analyzing the widespread and deep effects that a seemingly straightforward AI initiative can have on an organization led me to look for commonalities among disparate organizations forced to face similar pressures.

The observation that changed my thinking

What struck me was not that these organizations lacked expertise. They clearly didn't. But, the expertise remained fragmented. The organizations with which I worked (on AI in general or on specific AI initiatives in particular) all understood that the traditional IT evaluation process was clearly insufficient and inadequate.  As well, most seemed to gravitate towards additional discussions generally labeled as “governance”.  But, as these discussions got progressively fleshed out, each organization recognized the need to start bringing into the conversation several departments such as Legal, HR, Privacy, and others to address new questions, new concerns, new impacts (real or potential) for which AI was a catalyst.  This is where it dawned on me that these similarities were not haphazard events, but rather evidence that properly evaluating AI generally led to having new conversations.  It also led me to conclude that these organizations did have the expertise required to properly assess AI initiatives, but that the necessary conversations were not always taking place.  In fact, they rarely took place.

The more organizations I observed, the more convinced I became that they generally possessed the expertise needed to evaluate AI initiatives. What they often lacked were the conversations needed to connect that expertise.

That realization eventually led me to develop a different approach to preparing organizations for AI. In my next article, I will describe the framework that emerged from those observations—and how it helps organizations prepare for AI before implementation begins.