🤖〽️Prompt Engineering: Copilot Unabridged (Part 123: Why AI Is Not Bulletproof in Automation Scenarios)

Prompt Engineering Series

Prompt: "write a blogpost of 600 words on why AI is not bulletproof is automation scenarios"

Introduction

Artificial Intelligence (AI) has become the poster child of modern automation. From supply‑chain optimization to customer service chatbots to predictive maintenance in factories, it’s easy to assume that AI systems are precise, tireless, and - at least in theory - nearly infallible. But that assumption is exactly where organizations get into trouble. AI is powerful, yes, but bulletproof? Not even close. And understanding why it isn’t bulletproof is essential for anyone deploying automation in the next decade.

Let’s unpack the cracks beneath the shiny surface.

AI Learns From Data - And Data Is Messy

AI systems don’t understand the world; they understand patterns in data. And real‑world data is full of noise, bias, gaps, and contradictions.

• A model trained on historical hiring data may inherit past discrimination.
• A predictive maintenance system may fail if sensors degrade or environmental conditions shift.
• A customer‑service bot may misinterpret a request simply because the phrasing wasn’t in its training set. 

When the data is imperfect, the automation built on top of it inherits those imperfections. AI doesn’t magically 'fix' flawed data - it amplifies it.

Automation Assumes Stability, but the Real World Is Dynamic

Traditional automation works best in stable, predictable environments. AI‑driven automation is more flexible, but it still struggles when the world changes faster than the model can adapt.

Consider:

• Sudden market shifts
• New regulations
• Unexpected supply‑chain disruptions
• Novel user behaviors
• Rare edge‑case events

AI models trained on yesterday’s patterns can’t automatically understand tomorrow’s anomalies. Without continuous monitoring and retraining, automation becomes brittle.

AI Doesn’t 'Understand' - It Correlates

Even the most advanced AI systems don’t possess human‑level reasoning or contextual awareness. They operate on statistical correlations, not comprehension.

This leads to automation failures like:

• Misclassifying harmless anomalies as threats
• Failing to detect subtle but critical changes
• Producing confident but incorrect outputs
• Following rules literally when nuance is required

In high‑stakes environments - healthcare, finance, transportation - this lack of true understanding becomes a serious limitation.

Edge Cases Are the Achilles’ Heel

AI performs impressively on common scenarios but struggles with rare events. Unfortunately, automation systems often encounter exactly those rare events.

Examples include:

• A self‑driving car encountering an unusual road layout
• A fraud‑detection model missing a novel attack pattern
• A warehouse robot misinterpreting an unexpected obstacle

Humans excel at improvisation; AI does not. Automation breaks down when reality refuses to fit the training distribution.

Security Vulnerabilities Undermine Reliability

AI systems introduce new attack surfaces:

• Adversarial inputs can trick models with tiny, invisible perturbations.
• Data poisoning can corrupt training sets.
• Model inversion can leak sensitive information.
• Prompt manipulation can cause unintended behavior in language models.
• Automation built on AI can be manipulated in ways traditional systems never could.

Ethical and Governance Gaps Create Operational Risk

Even when AI works “correctly,” it may still cause harm if governance is weak.

Automation can:

• Reinforce bias
• Reduce transparency
• Remove human oversight
• Make decisions that lack accountability

Organizations often underestimate the social and regulatory risks of automated decision‑making. AI isn’t just a technical system - it’s a socio‑technical one.

Humans Are Still Part of the System

The biggest misconception about automation is that it removes humans. In reality, it changes the role of humans.

People must:

• Monitor AI outputs
• Intervene during failures
• Interpret ambiguous results
• Maintain and retrain models
• Handle exceptions and edge cases

If humans aren’t properly trained or workflows aren’t redesigned, automation becomes fragile.

The Bottom Line: AI Is Powerful, but Not Invincible

AI can supercharge automation, but it’s not a magic wand. It’s a tool - one that requires careful design, continuous oversight, and a deep understanding of its limitations. Organizations that treat AI as bulletproof will face costly failures. Those that treat it as a dynamic, fallible component of a broader ecosystem will unlock its real value.

Disclaimer: The whole text was generated by Copilot (under Windows 11) at the first attempt. This is just an experiment to evaluate feature's ability to answer standard general questions, independently on whether they are correctly or incorrectly posed. Moreover, the answers may reflect hallucinations and other types of inconsistent or incorrect reasoning.

Previous Post <<||>> Next Post

💫⚙️ERP Systems: Microsoft Dynamics 365's Vendor Invoices Processing (Setup Areas)

ERP Systems

Invoices can be set up to be imported automatically in Dynamics 365 for Finance and Operation (D365 F&O) via Invoice Capture (see process). Besides the Invoice Capture-related set up, there are several areas related to Vendor invoice's processing:

General Ledger

Accounts for automatic transactions need to be defined at least for the 'Vendor invoice rounding off' and 'Rounding variance', respectively for the 'Exchange rate gain' and 'Exchange rate loss', if this wasn't done already.

To facilitate accounts' reconciliation and reporting, one can enable a Batch transfer rule for the Source document type 'Vendor invoice' and thus one journal is used for each Vendor invoice. This also makes sure that the Description from Invoice will be taken over in the Journal, which facilitates accounts' reconciliation.

Subscription Billing

In case one needs to defer the amounts over a time interval (e.g. x months), as it's the case for prepayments, the integration with Subscription Billing on the Vendor side in F&O seems to work with minimal configuration. A Billing schedule is created for each Invoice distribution and can be modified or cancelled after the Invoice is posted, if needed.

If your organization uses Subscription Billing also for Accounts Receivables (AR), one might need to compromise on the setup because the same parameters are used for both modules. 

In what concerns accounts' reconciliation, there seems to be a 1:1 mapping between the Schedule line and the General Ledger (GL) posting, a table with the mapping between the records being available. One can build thus a Paginated report to display the mapping between AP and GL, if the infrastructure is in place (e.g. by building a data lakehouse/warehouse based on F&O, see post). 

Fixed Assets

One can enable the creation of Fixed assets by checking the "Create asset during product receipt or invoice posting" radio button in Fixed assets parameters.

The feature "create the fixed asset automatically during the time of invoice import" for PO-based Invoices (with "Create Fixed Asset" flag set at line level) doesn't seem to be supported yet.

Vendor Invoice Journals

Before Invoice Capture, Vendor invoice journals were helpful for posting summarized cost invoices that are not associated with POs (e.g. expenses for supplies or services). One can still use this approach, however if the line-based details are important, then it makes sense to use Pending vendor invoice with Service items or Procurement categories.  

A single Vendor invoice is created as one Vendor invoice journal. 

Organization Administration

When using workflows, setting the same language for all LEs can reduce the amount of redundant information maintained in the workflow(s), otherwise the texts need to be provided for each language. 

Default descriptions can be enabled for Purchase orders' invoice ledger and vendor to carry the same description entered Pending Vendor Invoices in Invoices and Journals. The functionality works also for Cost invoices.

Ideally, Invoice's description should have been maintained in Invoice capture and/or Microsoft should have provided a default description also for it. 

System

Unless there's a requirement to post manually the journals from subledger to GL, a batch transfer for subledger journals must be created for each LE.

One can enable email notifications for the users participating in workflows and use workflow delegations for the intervals the respective users are on leave. 

All users participating in the workflow must be available also as active employees. It makes sense to do latest when the integration goes Live. Moreover, the Users need to have the appropriate permissions for the roles they have in the process.

Vendor Invoice Automation

There's further functionality available under the 'Vendor Invoice Automation' label (see [1], [2]), though the following are the most important ones: 

• Automatically apply prepayments to vendor invoices
• Automatically submit imported invoices to the workflow system.
• Match product receipts to pending vendor invoice lines.

Recurring Vendor Invoice Templates

Microsoft introduced with 10.0.38 PU a new feature called 'Vendor invoice templates', which allows creating recurring vendor invoices without the need to enter all the vendor invoice information for each separate invoice (see [3]). There seems to be no information available whether any integration between this feature and Invoice Capture will be supported.

E-Invoicing

EU countries need to enforce the Directive 2014/55/EU for Vendor and Customer invoices. The directive requires that the electronic exchange of invoice documents between suppliers and buyers to occur over government-held third-party solutions. Each country has its own system(s) and regulations with different scope and timelines. Some countries have already requirements for 2024, respectively 2025 and there are similar projects in US and other countries. 

Even if Microsoft started this year (2023) to provide country-specific integrations for e-Invoicing, for the moment there seem no information available on how e-Invoicing will integrate with Invoice Capture.

Previous post <<||>> Next post

Resources:
[1] Microsoft Learn (2023) Vendor Invoice Automation (link)
[2]Microsoft Dynamics 365 (2023) The Future of Finance: Unlocking the Benefits of Accounts Payable Automation (link)
[3] Hylke Britstra (2023) Recurring vendor invoice templates (link)
[4] Dan Edwards (2023) AP Automation at DynamicsCon (link)