The Role of Artificial Intelligence in Information Management

A perspective from David Gould, Chief Customer Officer

It strikes me as paradoxical that businesses trust AI systems to guide life-or-death decisions, for example, when it is safe to make a lane change at 75 miles per hour. Yet they are reluctant to use the same technology to classify and manage information through its lifecycle. Artificial Intelligence is used almost every minute of every day, in nearly every smart device we own. 

For enterprise information management, AI is not a buzzword or a trendy topic. It’s a real technology with a real purpose, powering enterprise AI solutions. The emergence of machine learning, natural language processing, and other variants of Artificial Intelligence provide a significant advancement for:

• Discovering, classifying, and organizing data
• Consuming and retaining enterprise information
• Automating data governance
• Ensuring compliance with evolving regulations

AI is the key technology foundation enabling users to manage enterprise data at scale, take their data farther, and make it more dimensional than ever before.

The evolution of AI trust and explainability

My personal experience and perspective on AI has evolved significantly over the past 12 years. The biggest change is that results generated by AI are actually explainable.  You no longer are “required” to have PhDs in Bayesian Inference or Shannon Communication theories to understand what is happening behind the curtain. 

Today’s AI technology can demonstrate, and we can readily explain, how AI-driven systems process and produce results. There is a logical explanation available each and every time a result is provided, reinforcing trust in AI systems. Building trust in AI through responsible use has become a priority, ensuring transparency and explainability in enterprise applications.

Conversely, when I first helped bring the HP ControlPoint solution to market back in 2012, I remember how difficult – and uncomfortable –  it often was to explain the reasoning behind the result.  

Many still compare AI to a black box solution.

Like all technologies, AI is rapidly evolving and provides much more explainable reasoning behind the processing than ever before. There is a very noticeable lack of trust in AI systems to analyze, classify and manage information across the enterprise. In fact, there is a discernable level of apprehension, even disdain, among experienced and savvy information governance professionals, highlighting the ongoing need for building trust in AI systems through responsible, transparent implementation.

Why organizations hesitate to trust AI in information management

At EncompaaS, we have given this conundrum a great deal of thought – especially when it relates to how we are optimizing algorithms and providing “explanation” tools to effectively use AI to analyze and manage very large datasets. 

We have seen three core drivers that impact the adoption of AI-based solutions across the enterprise:

1. It’s a threat to my job

AI is a powerful enabling tool. Given the amount of daily data volume created and the need to “fact-check” the assigned manual classification, not even a huge team of individuals can keep pace. 

Instead of reclassifying manually, handle this with AI-driven information management solutions. Let the technology do that work for you. AI frees your time to focus on what is most important: creating and managing policy, enabling colleague productivity, and protecting your company’s brand reputation by discovering and protecting sensitive or crown jewel information.

2. Humans can do it better

The problem is that most large enterprises are creating upwards of 10,000 pieces of content an hour. 

It is physically impossible for those responsible for data management to physically review that size of information corpus without AI-enabled automation. Farming out file boxes of content to low-cost locations for manual review will undoubtedly produce a poor and inconsistent result. 

3. I don’t trust it

Customer trust in AI has been historically low due to poor explainability. Most of the focus was on the concept of confidence – how confident was the software that the document being returned was what the user expected.

However, accuracy – supported by recall – is a much better and more meaningful metric to evaluate the effectiveness of AI and building trust in AI systems. Unlike confidence scores, which have always been difficult to explain, accuracy and recall provide measurable, transparent insights that strengthen customer trust in AI.

Confidence alone has historically failed to build trust and reputation in AI, making it harder for potential buyers to feel assured in adopting AI-based auto-classification solutions. By prioritizing accuracy-driven AI governance, organisations can enhance trust in AI systems and drive broader enterprise adoption.

AI: The key to future-proofing enterprise information management

AI is the driving technology that will give information, data, and records managers more power, more control, and more extended uses of information than ever before.

At EncompaaS, we’re committed to building trust in AI through responsible use, helping organisations understand and embrace AI-powered transformation with confidence. Our goal is to make AI a trusted partner in enterprise information management, ensuring its adoption enhances governance, security, and data integrity at every stage.