Q 757. Imagine an AI system was asked to enforce a rule, policy, or guideline in your workplace. What’s one rule where you would fully trust AI to make the call — and one rule where you would never trust it? Why?

 

Note for website visitors -

• This platform hosts two weekly questions, one on Monday and the other on Thursday. 
• All previous questions can be found here: https://www.benchmarksixsigma.com/forum/lean-six-sigma-business-excellence-questions/. 
• To participate in the current question, please visit the forum homepage at https://www.benchmarksixsigma.com/forum/.
• The question will be open until Monday or Thursday at 5 PM Indian Standard Time, depending on the launch day. 
• Responses will not be visible until they are reviewed, and only non-plagiarised answers with less than 5-10% plagiarism will be considered for winner selection. 
• If you are unsure about plagiarism, please check your answer using a plagiarism checker tool such as https://smallseotools.com/plagiarism-checker/ before submitting. 
• All correct answers shall be published, and the top-rated answer will be displayed first. The author will receive an honourable mention in our Business Excellence dictionary at https://www.benchmarksixsigma.com/forum/business-excellence-dictionary-glossary/ along with the related term.

• Some people seem to be using AI platforms to find forum answers. This is a risky approach as AI responses are error-prone because our questions are application-oriented (they are never straightforward). Have a look at this funny example - https://www.benchmarksixsigma.com/forum/topic/39458-using-ai-to-respond-to-forum-questions/

• We also use an AI content detector at https://quillbot.com/ai-content-detector. Only answers with less than 45-50% AI-generated content will be considered for winner selection.

One rule where i would fully trust AI to make the call — i would trust AI to check compliance checks, simply because  AI can efficiently monitor systems and validate if compliance is met or not met.  AI can be used where rules are defined clearly to get an apt response.

 

One rule where i would never trust AI - AI may not be a suitable option to use where  human behavior is involved example disciplinary issues by employee, appraisal discussions etc.  This is best judged by relevant teams like HR who are trained to judge and understand situational behavior, emotions of employees, facial expressions etc. 

One rule I would fully trust AI to make a call on is managing and monitoring employee attendance and time tracking. This is because AI can effectively log working hours and detect absences. If escalations are noticed, it can send automated reminders to employees. In addition, since AI is unbiased, it can enforce the standards and rules for everybody without favoritism or neglect.

 

One rule I would never trust AI with would be taking care of workplace conflict resolution. More often than not, disputes between employees require emotions, personal experience, and observations which are beyond comprehension for AI. A human mediator is essential to listen, analyze, empathize, and make context-based decisions that ensure all angles are fairly considered. AI may misread sarcasm, tone, or intention, and cause an unfair or unfriendly outcome.

One rule I would fully trust AI to enforce:

“Timely submission of timesheets and project updates.”

AI can reliably track deadlines and send reminders or escalate delays based on predefined timelines. There’s no subjectivity involved, and the rule is based purely on observable data like timestamps, which AI handles well. It reduces the managerial burden and ensures better project governance.

 

One rule I would never trust AI to enforce:

“Evaluating employee performance or potential for promotion.”

Performance evaluations often involve soft skills, leadership potential, adaptability, and emotional intelligence — areas that are nuanced and context-driven. AI may rely heavily on quantitative metrics or biased datasets, risking unfair judgments. Human discretion is essential here to interpret intent, effort, and interpersonal dynamics.

There are certain things that AI excels at like logical reasoning, for e.g. shortlisting candidate profiles based on qualifications, invoice processing, monitoring inventory, IT helpdesk chatbots etc. On the contrary there are certain subjective questions where AI struggles and is prone to decision making mistakes just like human beings, which is also known as the “Linda problem”, for e.g. decision making on HR policies pertaining to employee's behavior, issuing warning letters to employees or even termination letters in worst case scenarios.

 

There is a stark distinction between traditional automation tools and AI, as AI is not dependent on fixed rule-based decision making, rather it learns from previous patterns and relationships to determine the best approach to address a given situation.

 

A detailed example of area best suited for AI deployment in corporates is pertaining to IT helpdesk. Say for e.g. as per company policy all queries should be answered/resolved within 24hrs. AI can be extremely useful to categorize the queries correctly, route it to the appropriate team and even provide resolution to tier one queries without any human intervention.

 

As AI is dependent on previous patterns to fetch the best answer, if previous patterns include bias, that bias will be carried forward by the AI in selecting the best approach. Therefore, subjective decision making like termination or issuing warning letter to an employee based on certain kind of employee behavior cannot be completely left up to AI. For e.g. There is a corporate rule pertaining to usage of social media sites in office. If an employee is found accessing social media site, AI cannot be fully trusted to pass a judgement in this scenario as there might be a lot of other factors to be considered before a warning letter can be issued.

One rule where I would trust the AI would be to create the audit/assessment calendar and allocate resources accordingly and the one where I would not trust the AI rule would be to interpret the legal or statutory and regulatory requirements, as mentioned in the contracts, because of various nuances associated to Stat & Reg or legal interpretations.

Tracking of completion of the mandatory trainings by all employees in my organisation can be entrusted to AI as they are completely driven by logic (like when they were last completed, whether an employees passed in the current assessment etc). At this point in time, I think with enough data, AI can enforce any rule or policy in an organization. With minimum data the outcome could be unpredictable and with more data accuracy of AI decisions will improve.

There are couple of examples I would like to suggest:

 

AI can be trusted - 

* Currently we have an AI deployed that addresses employee queries on HR policies at the first/ ground level. It provides guidelines and breaks down the complicated appearing policy terminologies in easy to understand language. AI can also re-direct/route random/ general queries regarding IT, HR, Admin an Facilities to the relevant team/s to accentuate timely and efficient query resolution

* AI can also be deployed for inventory management and ordering (from a pre-defined vendor) of freebies/ goodies that we distribute to recognize outstanding employees for implementing ideas and fusion of Ideas, also for continuous improvement projects that are deployed and implemented.

 

AI cannot be trusted - 

* On the other hand an AI intervention may not be trusted fully in the areas of discussing a feedback or the outcome of an audit that was conducted This is because the outcomes of the audits we conduct has severities (major, minor, observations and areas of improvement). There is a reasoning and dialogue that is required between the auditor and auditee to come to an agreement on the audit outcome and the way forward towards mitigation of the audit findings. At this moment an AI cannot be trusted to conduct this kind of complex task in an efficient manner as a human auditor would do. 

In the content moderation process that I am handling now, below are the rules that I would prefer AI to enforce and not prefer AI

 

Rules I will trust AI to enforce:

Content related enforcement:

AI can scan the content that is getting added, modified or deleted and highlight content that should not be permitted into the platform. This may be hate speech, cruelty, inappropriate content, etc.

 

Policy related enforcement:

Content moderation is governed by a set of policies. In these policies there are non gray areas where the AI could clearly call it out as an enforcement or violation. These cases can be handled by AI with high speed at higher accuracy than humans. 

 

Rules I will not trust AI to enforce:

Policy related enforcement:

There are certain gray areas in the policy which will require human intelligence. These pieces of validation are required to take the right call within the boundaries of the policies. Hence this is a task that I wont entrust to AI. 

 

Vidhya R

 

What’s one rule where you would fully trust AI to make the call?

Answer: Data backup and retention policy, where AI can make data-based decision for saving the files, archiving unused or old files and retaining data from backup as and when required.

 

one rule where you would never trust it? Why?

Answer: We would never trust AI to make Hiring and termination decisions as there are many impacting parameters that needs an human view to evaluate the situation and take decisions. It cannot be rule based though it can be done partially using AI like shortlisting resumes or documenting performance issues but not take decision if individual should be hired or terminated.

For any kind of data classification & data security related tasks I would trust AI completely as in such manual tasks there are more chances of Human error always and AI has proven the expertise and best results in these areas over the period with good data set and training models. 

 

Where I would not trust AI as much is in employee performance management & calibrations with stakeholder, these areas are quite subjective and requires human intervention, emotional intelligence and interpersonal conversion. In these areas there are high chances of AI failures only sharing the logical outcome or at the best basis the data on being the models are trained on which can be limited. While in these areas of work there are endless scenarios and situations which calls for unconventional approach and skills to deal with it.

AI Enforcement: Areas of Trust (Rule Where I Would Trust AI to Make the Call)

• Automated Workflows: Repetitive tasks with defined urgency/priority.
• Consistent & Data-Driven Tasks: Objective outcomes and clear parameters.
• Pattern Recognition (NLP): Identifying trends and categorizing information.

AI Enforcement: Areas Requiring Human Judgment (Rule Where I Would Never Trust AI to Make the Call)

• Contextual, Emotional, & Ethical Decisions: Nuance and human values are critical.
• Complex Technical Tasks: Deep understanding, hypothesis, and experimentation needed.
• Ambiguous/Incomplete Information: Requires interpretation and clarification.
• Non-Obvious Causes & Intuition: Leveraging experience and subtle cues.

 

If I were to depend on AI to implement a regulation within my workplace, it would certainly involve the monitoring and identification of overdue deadlines on various projects. AI is proficient in data analysis, identifying trends, and organizing schedules. It has the ability to manage SLAs, project deadlines, interdependencies, and to swiftly produce alerts without bias or delay. Unlike humans, it is not affected by multitasking nor does it tend to miss details due to cognitive fatigue. This capability would allow us to react more quickly, redistribute resources, and maintain project momentum. Certainly, I would feel more secure knowing there is a digital watchdog keeping timelines in check.

 

On the other hand, I would never assign to AI the responsibility of enforcing any issues that require an understanding of emotional subtleties, such as detecting inappropriate tone in emails or conversations. Language is full of context, culture, and emotion. What might appear harsh in one culture could be completely acceptable in another. I have witnessed instances where tone was misread even among colleagues, and AI, despite its progress, still lacks the emotional intelligence to truly grasp human intent. This is where human judgment, enhanced by empathy and communication, becomes vital.

 

In the end, AI can act as a useful assistant, but not every task is simple. Some requests require a human touch, rather than just an algorithm.

I am from US healthcare background, In my view i can trust on some of the repetitive tasks or data entry copy information from excel or PDF in to application with not much of complexity. and i dont trust on area where AI is the sole deciding maker on departments like patient payment plans or patient calling where you need a human judgement to take decision 

It was interesting to read the different perspectives and answers to the one rule that you can completely trust with AI

 

Best answer has been given by Swarandeep Kaur Juneja. Well done.

 

Answers from Amit Suri, Vikas Choudhary are a must read.