The age of employees’ sneaking AI into the workplacespan “font-weight 400 ;”>’ made it clear that transformative AI is best used from the beginning. Staff in various roles, from marketing, sales, and operations, are increasingly introducing AI tools to their work. They do this to automate processes, streamline tasks, and simplify everyday tasks.
Microsoftspan size=”font-weight 400 ;”>”s Work Trend Index revealed that employees were comfortable with AI and eager to implement it in their workplaces regardless of any formal policies. This trend, dubbed “Bring Your Own AI”, signals that individuals are able to identify processes that can be automated faster than leaders.
In this context, “smuggling AI in work” is less of a subversion than it is a sign that workers are willing to push boundaries and embrace new opportunities. As a testament to this, millions of people in the UK already use AI. Nearly three quarters of them report a significant increase in productivity.
The remaining organisations who use a top-down rigid approach to AI exploration are in stark contrast with this bottom-up method. What happens if the organisation as a whole is not involved in exploring a new technology or tool? Integration is slow and cumbersome, with a lack of desire to progress along with the tool.
The adoption of AI by employees is not without its challenges. Consistency in usage can lead to data silos, which are a security risk. If teams are not properly guided or supported, a hands-off approach can lead to confusion and even disillusionment.
How can we find a balance between caution and innovation?
Leaders must create clear frameworks that encourage experimentation to avoid these pitfalls. It is important to establish policies that safeguard data security and integrity, promote upskilling programs for AI fluency and ensure all necessary technology infrastructure.
Maintaining a comprehensive context is essential for consistency and quality as AI tools evolve from simple tasks to more complex processes that require multiple steps. Businesses risk AI tools not working together, either complicating tasks to fill in context gaps, or settling for less-than-ideal outcomes, without a unified data and connection view. Graph technology can be a real asset in this case by connecting information across the organization.
This can be supported by a graph database approach, as knowledge graph technology will reveal the connections in an organisation’s data. This allows for a holistic view of AI applications and insights. By revealing the relationships between data entities in graph databases, they help to break down silos which threaten to derail bottom up AI initiatives. It is easier for employees who need to test new ideas because the data needed is readily available and reliable.
Graph solutions allow each user to discover hidden patterns and connections within their “own data universe”, surfacing the context most relevant for a specific task. Over time, this can lead to meaningful collaboration as teams seamlessly share insights, while governance and security protocols of the organization track data flows. Graph technology allows for experimentation at the grassroots while also providing a framework for enterprise-wide AI adoption.
If done correctly, grassroots AI adoption is a strategic asset that can transform entire organisations. Klarna is a good example. Its co-founder, Sebastian Siemiatkowski explained that the company decided to “explore the potential of AI/LLMs and be open to testing everything that seemed to trend” and that it “allowed employees to pursue ideas organically instead of following management directions on what they should build.” This led to new innovations and agile problem solving. Siemiatkowsi says that companies should “go where the puck’s going” and anticipate the future, embracing new approaches and tools while momentum is there. Klarna’s use of graph-based technologies demonstrates how companies can scale their employees’ AI driven ideas while maintaining a contextualised, shared view of data.
Your company can embed AI benefits into its culture if you understand that the real transformation starts when the people closest to the challenges create the solutions.
In the end, AI implementation that is meaningful does not have to be imposed from above. It thrives when workers can learn, experiment and integrate AI in areas where it adds real value. This is a lesson about trust, agility and realizing that the most profound impacts of technology are created from the bottom-up: by one employee, with one idea and one experiment.
Andreas Kollegger Gen AI Innovation Lead at Neo4j
The first time this post appeared was on Human Resources News.
