Risks of Using LLMs in Your Business – What Does OWASP Have to Say?

Jesper Grode | April 10, 2024 | 4 minute read

Ready to learn more about MDM prompt libraries?

Get in touch

Risks of Using LLMs in Your Business – What Does OWASP Have to Say?

Master Data Management Blog by Stibo Systems logo
| 4 minute read
April 10 2024
Risks of Using LLMs in Your Business – What Does OWASP Have to Say?
7:17

Powerful artificial intelligence (AI) tools like Large Language Models (LLMs) have been making headlines as companies consider how the technology can be used to drive more efficient operations and better customer experiences. One very promising LLM use case is in product experience management. But as companies accelerate to integrate LLMs into their business applications, there are some risks that should be considered along the way.   


Benefits of using LLMs to enrich the product experience 

With more of the shopping experience occurring online, delivering a seamless and engaging digital experience has become table stakes. Product experience management helps ensure users have access to detailed and accurate product information, along with rich content and imagery, that communicates a product’s value, benefits, footprint and more.

Delivering better product experiences can help generate future revenue while also mitigating the revenue loss and logistic headaches associated with product returns by ensuring consumers have all the information needed to make the right product choice. 

By using LLMs as part of a master data management (MDM) prompt library, companies can create rich, compelling product descriptions in a faster and more efficient way for various segments and channels. 

In this LLM use case, it’s important that the prompt library is part of the foundational MDM design, which integrates advanced security measures, flexible prompt crafting, and thorough content review processes. Without this integrated approach, companies open themselves up to the risks of prompt injections, insecure output handling and other vulnerabilities. 
 

Potential risks of using LLMs and how to safeguard against them 

The Open Worldwide Application Security Project (OWASP) recently released its list of top 10 vulnerabilities for LLMs, which include prompt injections; insecure output handling; training data poisoning; denial of service; supply chain; permission issues; data leakage; excessive agency; overreliance; and insecure plugins.

These vulnerabilities have the potential to introduce several risks to your business, ranging from security issues, system failures and service issues to compromised data integrity from the generation of misinformation, inappropriate or biased content. 

Being aware of the risks of LLMs gives you the opportunity to turn those vulnerabilities into strengths by ensuring your AI initiatives are governed and grounded in strong security principles. 

Let’s look at a few of these vulnerabilities – in the context of product experience management – and how using an MDM prompt library can help mitigate the risks:

 
  • Prompt injections: When using LLMs to enrich the data used in product descriptions, indirect or direct prompt injections could result in product descriptions with malicious content, or poor product descriptions that negatively impact conversion rates. For product descriptions created in bulk, this can be a costly mistake, as it requires a manual clean-up effort to remedy. 

    An MDM prompt library mitigates these risks, as the prompts are crafted in advance by prompt engineers. This ensures built-in functions that check for malicious content in prompts, effectively preventing prompt injection. 
  • Insecure output handling: This vulnerability can occur when an LLM output is accepted by downstream systems without any review or interaction to confirm the content is factual and meets the objectives.

    Adhering to the principles of Responsible AI and building human-centric AI systems can eliminate the creation of poor-quality content or security concerns. With this approach, product descriptions created using an MDM prompt library are then passed on as suggestions to content reviewers, who approve the content for usage. Although the content is AI-generated, it incorporates a “human-in-the-loop” protocol.

    Be sure to use an MDM with advanced and configurable workflows, which make it possible for the prompt library capabilities to be embedded into any content authoring process across the business. 
  • Training data poisoning: In a product experience management scenario, relying on the foundation models to generate product descriptions implies a direct dependency on any bias introduced into the models when they were trained.

    Businesses have no influence on how models are trained, which makes it even more important that the use of models is done with careful consideration of any bias that the models may have. Bias takes many forms, such as selection bias, confirmation bias, measurement bias, and stereotyping bias.

    With rich data sets from MDM, models can be fine-tuned to provide a better understanding of the businesses’ product offerings. In clothing manufacturing, for example, information about fabric composition, washing instructions, etc., could be relevant to use when fine-tuning a model to provide better product descriptions.

    An MDM prompt library that is built to include a human-in-the-loop protocol guarantees that the outputs are validated and scrutinized for bias. Additionally, using the data sets from the entire product data set in the MDM for fine-tuning an LLM will further minimize the risk of bias. The master data governance processes allow users to enrich data to the highest or desired quality standard, which in return can be fed into the LLM, as part of the training or grounding process.
  • Sensitive information disclosure: LLMs may inadvertently reveal confidential data in its responses, leading to unauthorized data access, privacy violations and security breaches. It’s crucial to implement data sanitization and strict user policies to mitigate this. Through strong data governance policies and the possibility to encrypt API keys, and through careful design of data access through user role and responsibility setup, these risks are effectively mitigated.

While these are just a few examples, following are some other recommendations to consider when using LLMs to create lasting consumer engagement through creative and crafted product experience creation:

  • Use a private LLM (i.e., Open AI on Microsoft Azure) to ensure data privacy
  • Create alerts in case of service unavailability or long response times
  • Configure different services in different regions as a fallback
  • Use an open platform MDM that enables connections to LLMs via REST API


By embedding the principles of Responsible AI and human-in-the-loop functionalities, the MDM prompt library enhances the quality and relevance of AI-generated content. It ensures that the output not only meets but exceeds the highest standards of quality and safety, all while boosting operational efficiency and revenue opportunities and safeguarding against AI-related threats.

Look for an MDM partner that’s committed to data integrity and will make your company’s security, efficiency, and innovation a priority, turning the use of LLMs into a strategic advantage while carefully observing and mitigating the risks associated herewith.

 
 

Master Data Management Blog by Stibo Systems logo

Decades of experience within master data management, technologies, people and processes has led Jesper into his current role, heading Stibo Systems' innovation efforts. He has a particular focus on multidomain MDM, augmented MDM and technology adoption. Being responsible for company-wide strategic initiatives on product innovations, he is constantly seeking to increase the value of product offerings to customers and partners. Jesper comes from prior roles as Product Strategy Director, Section Head R&D, Director Professional Services, and Associate Professor at a Danish university.

Discover Blogs by Topic

  • MDM strategy
  • Retail and distribution
  • Data governance
  • See more
  • Customer and party data
  • Manufacturing
  • Product data and PIM
  • Data quality
  • AI and machine learning
  • Supplier data
  • CPG
  • Financial services
  • Sustainability
  • GDPR
  • Location data
  • PDX Syndication
  • Customer Experience
  • Product Experience Data Cloud
  • Cloud
  • Microsoft Azure
  • Product Onboarding

Gen Z: Seeking Excitement Beyond Amazon

12/11/24

A Modern Guide to Data Quality Monitoring: Best Practices

12/10/24

CDP and MDM: Complementary Forces for Enhancing Customer Experiences

12/10/24

Using Machine Learning and MDM CBAM for Sustainability Compliance

12/3/24

How to Implement Master Data Management: Steps and Challenges

11/26/24

AAPEX and SEMA: The Automotive Aftermarket Industry’s Mega-Showcase

11/25/24

5 Key Trends in Product Experience Management

11/20/24

Building the Future of Construction with AI and MDM

11/19/24

Solving Retail Data Fragmentation: The Key to Consistent Customer Journeys

11/14/24

Live Shopping: How to Leverage Product Information for Maximum Impact

10/22/24

Why Data Accuracy Matters for CPG Brands

10/16/24

Why Choose a Cloud-Based Data Solution: On-Premise vs. Cloud

10/15/24

How to Use Customer Data Modeling

10/10/24

Navigating Change: Engaging Business Users in Successful Change Management

9/20/24

What is Digital Asset Management?

9/11/24

How to Improve Your Data Management

9/3/24

The Future of Master Data Management: Trends in 2023-2025

9/1/24

Digital Transformation in the CPG Industry

8/30/24

5 CPG Industry Trends and Opportunities for 2024-2025

8/29/24

What is the difference between CPG and FMCG?

8/27/24

Responsible AI relies on data governance

8/27/24

6 Features of an Effective Master Data Management Solution

8/15/24

Great Data Minds: The Unsung Heros Behind Effective Data Management

8/13/24

A Data Monetization Strategy - Get More Value from Your Master Data

8/6/24

Introducing the Master Data Management Maturity Model

8/4/24

What is Augmented Data Management? (ADM)

7/31/24

Data Migration to SAP S/4HANA ERP - The Fast and Safe Approach with MDM

7/30/24

GDPR Data Governance and Data Protection, a Match Made in Heaven?

7/17/24

The Difference Between Master Data and Metadata

5/26/24

Master Data Management Roles and Responsibilities

5/20/24

8 Best Practices for Customer Master Data Management

5/16/24

What Is Master Data Governance – And Why Do You Need It?

5/12/24

Guide: Deliver flawless rich content experiences with master data governance

4/11/24

Risks of Using LLMs in Your Business – What Does OWASP Have to Say?

4/10/24

Guide: How to comply with industry standards using master data governance

4/9/24

Digital Product Passports - A Data Management Challenge

4/8/24

Guide: Get enterprise data enrichment right with master data governance

4/2/24

Guide: Getting enterprise data modelling right with master data governance

4/2/24

Guide: Improving your data quality with master data governance

4/2/24

Data Governance Trends 2024

1/30/24

NRF 2024 Recap: In the AI era, better data can make all the difference

1/19/24

Building Supply Chain Resilience: Strategies & Examples

12/19/23

How Master Data Management Can Enhance Your ERP Solution

12/14/23

Shedding Light on Climate Accountability and Traceability in Retail

11/29/23

What is Smart Manufacturing and Why Does it Matter?

10/11/23

Future Proof Your Retail Business with Composable Commerce

10/9/23

5 Common Reasons Why Manufacturers Fail at Digital Transformation

10/5/23

How to Digitally Transform a Restaurant Chain

9/29/23

Three Benefits of Moving to Headless Commerce and the Role of a Modern PIM

9/14/23

12 Steps to a Successful Omnichannel and Unified Commerce

7/6/23

CGF Global Summit 2023: Unlock Sustainable Growth With Collaboration and Innovation

7/5/23

Navigating the Current Challenges of Supply Chain Management

6/28/23

Product Data Management during Mergers and Acquisitions

4/6/23

A Complete Master Data Management Glossary

3/14/23

4 Ways to Reduce Ecommerce Returns

3/8/23

Asset Data Governance is Central for Asset Management

3/1/23

4 Common Master Data Management Implementation Styles

2/21/23

How to Leverage Internet of Things with Master Data Management

2/14/23

Manufacturing Trends and Insights in 2023-2025

2/14/23

Sustainability in Retail Needs Governed Data

2/13/23

NRF 2023: Retail Turns to AI and Automation to Increase Efficiencies

1/20/23

5 Key Manufacturing Challenges in 2023

1/16/23

What is a Golden Customer Record in Master Data Management?

1/9/23

Innovation in Retail

1/4/23

Life Cycle Assessment Scoring for Food Products

11/21/22

Retail of the Future

11/14/22

Omnichannel Strategies for Retail

11/7/22

Hyper-Personalized Customer Experiences Need Multidomain MDM

11/5/22

What is Omnichannel Retailing and What is the Role of Data Management?

10/25/22

Most Common ISO Standards in the Manufacturing Industry

10/18/22

How to Get Started with Master Data Management: 5 Steps to Consider

10/17/22

What is Supply Chain Analytics and Why It's Important

10/12/22

What is Data Quality and Why It's Important

10/12/22

An Introductory Guide: What is Data Intelligence?

10/1/22

Revolutionizing Manufacturing: 5 Must-Have SaaS Systems for Success

9/15/22

An Introductory Guide to Supplier Compliance

9/7/22

What is Application Data Management and How Does It Differ From MDM?

8/29/22

Digital Transformation in the Manufacturing Industry

8/25/22

Master Data Management Framework: Get Set for Success

8/17/22

Discover the Value of Your Data: Master Data Management KPIs & Metrics

8/15/22

Supplier Self-Service: Everything You Need to Know

6/15/22

Omnichannel vs. Multichannel: What’s the Difference?

6/14/22

Create a Culture of Data Transparency - Begin with a Solid Foundation

6/10/22

The 5 Biggest Retail Trends for 2023-2025

5/31/22

What is a Location Intelligence?

5/31/22

Omnichannel Customer Experience: The Ultimate Guide

5/30/22

Location Analytics – All You Need to Know

5/26/22

Omnichannel Commerce: Creating a Seamless Shopping Experience

5/24/22

Top 4 Data Management Trends in the Insurance Industry

5/11/22

What is Supply Chain Visibility and Why It's Important

5/1/22

The Ultimate Guide to Data Transparency

4/21/22

How Manufacturers Can Shift to Product as a Service Offerings

4/20/22

How to Check Your Enterprise Data Foundation

4/16/22

An Introductory Guide to Manufacturing Compliance

4/14/22

Multidomain MDM vs. Multiple Domain MDM

3/31/22

Making Master Data Accessible: What is Data as a Service (DaaS)?

3/29/22

How to Build a Successful Data Governance Strategy

3/23/22

What is Unified Commerce? Key Advantages & Best Practices

3/22/22

How to Choose the Right Data Quality Tool?

3/22/22

What is a data domain? Meaning & examples

3/21/22

6 Best Practices for Data Governance

3/17/22

5 Advantages of a Master Data Management System

3/16/22

A Unified Customer View: What Is It and Why You Need It

3/9/22

Supply Chain Challenges in the CPG Industry

2/24/22

The Best Data Governance Tools You Need to Know About

2/17/22

Top 5 Most Common Data Quality Issues

2/14/22

What Is Synthetic Data and Why It Needs Master Data Management

2/10/22

What is Cloud Master Data Management?

2/8/22

How to Implement Data Governance

2/7/22

Build vs. Buy Master Data Management Software

1/28/22

Why is Data Governance Important?

1/27/22

Five Reasons Your Data Governance Initiative Could Fail

1/24/22

How to Turn Your Data Silos Into Zones of Insight

1/21/22

How to Improve Supplier Experience Management

1/16/22

​​How to Improve Supplier Onboarding

1/16/22

How to Enable a Single Source of Truth with Master Data Management

1/13/22

What is a Data Quality Framework?

1/11/22

How to Measure the ROI of Master Data Management

1/11/22

What is Manufacturing-as-a-Service (MaaS)?

1/7/22

The Ultimate Guide to Building a Data Governance Framework

1/4/22

Master Data Management Tools - and Why You Need Them

12/20/21

The Dynamic Duo of Data Security and Data Governance

12/20/21

How to Choose the Right Supplier Management Solution

12/20/21

How Data Transparency Enables Sustainable Retailing

12/6/21

What is Supplier Performance Management?

12/1/21

What is Party Data? All You Need to Know About Party Data Management

11/28/21

What is Data Compliance? An Introductory Guide

11/18/21

How to Create a Marketing Center of Excellence

11/14/21

The Complete Guide: How to Get a 360° Customer View

11/7/21

How Location Data Adds Value to Master Data Projects

10/29/21

How Marketers Should Prepare for the 2023 Holiday Shopping Season

10/26/21

What is Supplier Lifecycle Management?

10/19/21

What is a Data Mesh? A Simple Introduction

10/15/21

How to Build a Master Data Management Strategy

9/26/21

10 Signs You Need a Master Data Management Platform

9/2/21

What Vendor Data Is and Why It Matters to Manufacturers

8/31/21

3 Reasons High-Quality Supplier Data Can Benefit Any Organization

8/25/21

4 Trends in the Automotive Industry

8/11/21

What is Reference Data and Reference Data Management?

8/9/21

What Obstacles Are Impacting the Global Retail Recovery?

8/2/21

GDPR as a Catalyst for Effective Data Governance

7/25/21

All You Need to Know About Supplier Information Management

7/21/21

5 Tips for Driving a Centralized Data Management Strategy

7/3/21

Welcome to the Decade of Transparency

5/26/21

How to Become a Customer-Obsessed Brand

5/12/21

How to Create a Master Data Management Roadmap in Five Steps

4/27/21

What is a Data Catalog? Definition and Benefits

4/13/21

How to Improve the Retail Customer Experience with Data Management

4/8/21

How to Choose the Right Master Data Management Solution

3/29/21

Business Intelligence and Analytics: What's the Difference?

3/25/21

Spending too much on Big Data? Try Small Data and MDM

3/24/21

What is a Data Lake? Everything You Need to Know

3/21/21

How to Extract More Value from Your Data

3/17/21

Are you making decisions based on bad HCO/HCP information?

2/24/21

Why Master Data Cleansing is Important to CPG Brands

1/20/21

CRM 2.0 – It All Starts With Master Data Management

12/19/20

5 Trends in Telecom that Rely on Transparency of Master Data

12/15/20

10 Data Management Trends in Financial Services

11/19/20

Seasonal Marketing Campaigns: What Is It and Why Is It Important?

11/8/20

What Is a Data Fabric and Why Do You Need It?

10/29/20

Transparent Product Information in Pharmaceutical Manufacturing

10/14/20

How to Improve Back-End Systems Using Master Data Management

9/19/20

8 Benefits of Transparent Product Information for Medical Devices

9/1/20

How Retailers Can Increase Online Sales in 2023

8/23/20

Master Data Management (MDM) & Big Data

8/14/20

Key Benefits of Knowing Your Customers

8/9/20

Women in Master Data: Kelly Amavisca, Ferguson

8/5/20

Customer Data in Corporate Banking Reveal New Opportunities

7/21/20

How to Analyze Customer Data With Customer Master Data Management

7/21/20

How to Improve Your 2023 Black Friday Sales in 5 Steps

7/18/20

4 Ways Product Information Management (PIM) Improves the Customer Experience

7/18/20

How to Estimate the ROI of Your Customer Data

7/1/20

Women in Master Data: Rebecca Chamberlain, M&S

6/24/20

How to Personalise Insurance Solutions with MDM

6/17/20

How to Democratize Your Data

6/3/20

How to Get Buy-In for a Master Data Management Solution

5/25/20

How CPG Brands Manage the Impact of Covid-19 in a Post-Pandemic World

5/18/20

5 Steps to Improve Your Data Syndication

5/7/20

Marketing Data Quality: Why Is It Important and How to Get Started

3/26/20

Panic Buying: Navigating Long-term Implications and Uncertainty

3/24/20

Women in Master Data: Ditte Brix, IMPACT

2/20/20

Get More Value From Your CRM With Customer Master Data Management

2/17/20

Women in Master Data: Nagashree Devadas, Stibo Systems

2/4/20

How to Create Direct-to-Consumer (D2C) Success for CPG Brands

1/3/20

Women in Master Data: Anna Schéle, Ahlsell

10/25/19

Women in Master Data: Morgan Lawrence, Infoverity

9/26/19

Women in Master Data: Sara Friberg, Acando (Part of CGI)

9/13/19

Improving Product Setup Processes Enhances Superior Experiences

8/21/19

How to Improve Your Product's Time to Market With PDX Syndication

7/18/19

8 Tips For Pricing Automation In The Aftermarket

6/1/19

How to Drive Innovation With Master Data Management

3/15/19

Discover PDX Syndication to Launch New Products with Speed

2/27/19

How to Benefit from Product Data Management

2/20/19

What is a Product Backlog and How to Avoid It

2/13/19

How to Get Rid of Customer Duplicates

2/7/19

4 Types of IT Systems That Should Be Sunsetted

1/3/19

How to Reduce Time-to-Market with Master Data Management

10/28/18

How to Start Taking Advantage of Your Data

9/12/18

6 Signs You Have a Potential GDPR Problem

8/16/18

GDPR: The DOs and DON’Ts of Personal Data

6/13/18