To owners and managers of small and medium-sized enterprises: have you ever felt a “perception gap” between your company’s rules and on-the-ground reality? Particularly regarding new ways of working like side jobs and multiple job holdings, differing internal perceptions are a challenge for many companies.
A recently published cross-industry survey on side job compliance highlighted this very “perception gap.” Meanwhile, another news report covered movements to use AI to “design” governance and “implement” it within systems. At first glance, these two stories seem unrelated. But when we connect them, we can see a future form of governance: a new practice of “predicting risks, designing for them, and embedding them with tools.”
The “Three-Party Gap” Reveals “Design Flaws” in Rules
According to a report by Mapion, a survey revealed a significant gap in the perception of side job issues among three parties: “executives,” “management departments,” and “employees.” While executives are optimistic, believing “no major problems have occurred,” on-the-ground management departments “sense potential risks,” and employees “do not fully understand the rules themselves.”
This is a typical governance “design flaw” seen in many SMEs. Rules (like side job policies) are created, but they are often merely “lists of prohibitions.” ① The purpose of the rule, ② the acceptable level of risk, and ③ the decision-making process when questions arise are not permeating all levels of the organization. As a result, the rules become shelf decorations, drifting apart from actual practice.
Governance-Aware AI: Transforming Rules into “Working Mechanisms”
Now, let’s consider the potential shown by the second piece of news. A PR TIMES article introduced “governance-aware AI-driven development” utilizing Anthropic’s “Claude.” This initiative involves AI automatically checking development processes against pre-set governance policies (security, privacy, compliance, etc.), warning of deviations, or suggesting appropriate code.
What happens if we apply this concept to areas like HR and labor management, such as the “side job gap”? For example, when an employee submits a side job application, an AI could instantly analyze the form’s content and provide feedback like: “These conditions carry a medium risk (level 50) of conflicting with non-compete obligations. Suggested adjustments A and B are provided.” Or, for the management department, a tool could learn from past application cases and their approval/denial reasoning to support consistent decisions in similar cases. Such tools hold the potential to transform static “regulations” into dynamic, interactive “decision-support systems.”
Designing Risk on a “1-99” Scale and Implementing it with AI
This media outlet consistently argues that risk should be designed not as “0 or 100,” but as a “continuous scale from 1 to 99.” For side jobs, this means neither “complete prohibition (risk 0 but opportunity cost 100)” nor “complete freedom (risk 100).” Instead, individually assess risks like “information leakage,” “labor provision,” and “competition,” and set acceptable levels for each (e.g., competition risk tolerance below 10, information leakage risk tolerance below 30).
This “design of risk tolerance” is the core work of management. Translating and implementing that design into systems, including AI, can be considered the role of the modern management department. AI can act as training wheels, reducing the “judgment fluctuations due to fatigue” of managers facing countless applications and aiding in consistent decisions aligned with the risk tolerance designed by executives.
Three Actions SMEs Can Start Today
Introducing advanced AI tools might seem like a high hurdle. However, with the right mindset, you can start tomorrow.
1. Visualize the Reality of the “Three-Party Gap”
First, measure your company’s “perception gap.” Using your side job policy as a subject, conduct a simple survey with executives, department managers, and general employees (a sample is fine). Example questions:
- “What do you think is the purpose of our company’s side job policy?” (open-ended)
- “Please give one specific example of a side job that would be considered competitive.” (open-ended)
- “What is the most important criterion when judging a side job application?” (multiple choice)
Simply comparing the answers will reveal where the interpretation of the rules is fragmented. This is understanding the “As-Is” (current state).
2. “Design” and Document Your Risk Tolerance
Next, the management team should design the “To-Be” (desired state). Discuss “which risks related to side jobs our company is willing to tolerate, and to what extent,” and document it as follows:
Example: “Our company permits side jobs under certain conditions to encourage employee career development and acquisition of new knowledge. However, we consider the following three points as significant risks: ① Impairment of primary job performance (reduction in working hours/concentration), ② Risk of leaking our confidential information, ③ Conflict of interest due to direct competition—we strive to avoid ③ as much as possible (acceptable risk level below 10). When making judgments, we place the highest importance on the transparency of the application (job content, hours, involved parties).”
This document becomes the “blueprint” for all future decisions. Simply adding it as a preamble to your policy can be effective.
3. Make Decisions “Visible” and Create a Feedback Loop
Finally, accumulate decisions based on this blueprint and use them for improvement. Each time a side job application arrives, the management department should rate each of the three risks above as low, medium, or high, and briefly record the decision (approval, conditional approval, denial) and its reason. Review these records regularly (e.g., quarterly) to examine “whether decisions align with the blueprint” and “whether the blueprint itself needs adjustment.”
This “record and review” process can be started with any tool, be it Excel or SharePoint. If you envision eventually incorporating AI analysis or support here, that is a solid first step toward “governance-aware” management.
Tools Don’t Replace Thinking. They Extend It.
AI and digital tools do not replace the essential thinking of executives: “designing risk.” Rather, they are “extension devices” to permeate that thinking throughout the entire organization and execute it consistently and sustainably. The “perception gap” revealed by the side job survey is a problem of thinking not permeating the organization. The news about governance-aware AI shows the potential for technology to support that permeation.
First, check if your company’s rules are more than just a “list of prohibitions” and if they contain the executive’s intent of “designing risk tolerance.” Then, digitize the process of communicating, implementing, and verifying that intent as much as possible. Accumulating these efforts will shape a next-generation SME governance that is agile, robust, and rivals that of large corporations.
It can be said that the evolution of governance is no longer about creating thick rulebooks, but a competition in the “technology of implementation”—how smartly we can disseminate management’s intent to every corner of the organization.
Comments