Is The Law Of The Instrument Hindering Enterprise AI Investments?

Forbes

30-06-2025

Two recent studies found that many enterprises are currently experiencing mixed results with their AI initiatives and investments. For example, Accenture examined over 2,000 generative AI projects and consulted more than 3,000 C-level executives, finding that only 36% stated they have scaled their generative AI solutions, while only 13% of executives report having created 'significant enterprise-level value.' Meanwhile, IBM surveyed more than 2,000 CEOs globally and found that only 16% have successfully scaled their AI initiatives across the enterprise, and only 25% report that their AI initiatives have delivered the expected ROI over the past few years.
The studies also suggest that there is a range of things that organizations have to get right if they are to be successful in harnessing the potential of generative AI. These include leadership alignment, enterprise strategy, data cleanliness and availability, the need for a modern technological infrastructure, the right internal skills and capabilities and the ability to manage large-scale change.
However, the Accenture study also found that organizations with 'leaders who deeply understand generative AI' are six times more likely to achieve enterprise-level value from their AI investments. This finding, combined with all of the hype currently circulating in the market about generative AI, the capabilities of LLMs, and, now, agentic AI, led me to wonder if there is possibly another factor at play here: the Law of the Instrument.
If all you have is a hammer, then everything looks like a nail.
The Law of the Instrument is a cognitive bias that occurs when we acquire a new skill or tool, leading us to over-rely on it and try to use it everywhere, even if it might not be the most suitable or effective solution.
This law is widely attributed to Abraham Maslow, who in 1966 wrote in his book The Psychology of Science: A Reconnaissance, "It is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail."
These days, it is often shortened to something along the lines of, 'If all you have is a hammer, then everything looks like a nail.'
So, is it possible that generative AI is becoming the proverbial hammer?
Perhaps.
Prayag Narula, CEO and co-founder of HeyMarvin, an AI-native customer feedback repository, believes that "Companies often approach AI like it's a one-size-fits-all solution'. Instead, he believes that "the better approach is to stop expecting magic. Start with systematic listening, pattern identification, and then match the right tools to specific problems.'
This type of approach, which leverages what generative AI excels at — namely, identifying relationships within multimodal datasets and then acting upon those insights — is very much aligned with the approach that Five9, a leading cloud contact centre and AI services provider, is taking with its clients.
According to Five9's CPO, Ajay Awatramani, they are advising their clients to be led by their data rather than by the tools at their disposal. As a result, whenever they engage with a client, the first thing they do is to listen via their tools, resources, and prompt engines to the massive multimodal datasets that their clients have of the interactions they have with their customers.
Beginning this way, Awatramani says, allows them to identify patterns in the data, such as common customer issues, reasons behind customer calls, or areas of frustration with their overall service experience. Once identified, some of these patterns lend themselves to self-service solutions, some to agent-assist solutions, while others may point to issues that go beyond the scope of the customer service leader, requiring a more strategic discussion with other parts of the business because, for example, the pattern that has been identified indicates the need for something like a new digital front end.
This seems like an eminently sensible approach, as it leverages one of the key strengths of generative AI while also focusing on understanding the problem space before considering possible solutions.
Moreover, this type of approach is likely to be easier to scale and to deliver a clear RoI as it ensures that companies avoid the Law of the Instrument and pick the right tool for the right job.