Training, not restraining: the only viable response to shadow AI

The use of generative AI in the workplace is growing at a dizzying pace, and what’s more, often without the knowledge or approval of IT departments or management.

Dubbed shadow AI, this is something we saw more than a decade ago with shadow IT: the use of technology by some employees independently to circumvent the limitations and restrictions imposed by the IT department, as I commented on in 2012 in this article (in Spanish) about the use of Dropbox in corporate environments back then.

Generative AI is unstoppable. Recent studies indicate that a significant proportion of employees use it without official approval from their organizations, with up to 50% of employees using unauthorized AI technology in their day-to-day tasks. This practice, although likely to improve productivity, can also pose serious compliance and security risks.

The solution lies not in cracking down on it, but in training the workforce, an approach that can give companies a competitive advantage. Teaching people do go beyond mere prompting, so that they also understand the logic and implications of AI, as well as its advanced use, is essential to mitigate risks and take advantage of their benefits.

Rather than futilely trying to prevent their employees from using unauthorized AI…