AI Risk Management Frameworks and Competitive Advantages for Small Businesses

It is now clear that organisations have significantly increased adoption of technology with AI components, specifically Generative AI components. While a lot of studies have focused on the measurable benefits of the increased adoption of Generative AI (GenAI), one McKinsey study also looks at the increase in the risks that are associated with GenAI and the slow adoption of risk mitigation practices across industries.

Specifically, in working with marketing and ad agency businesses, notable brands have adopted assessment mechanisms to vet the agencies they work with to check for AI Trustworthiness and ensure the risks associated with AI are optimized. Such assessments have now become the norm in RFI and RFP processes of brands.  

However, small and mid-sized agencies are slow to adopt a robust AI governance program that manages to score well on such assessments. As a result, agencies that have spent some time and efforts to put in place Risk Management Frameworks (RMFs) and optimize for the risks associated with GenAI have a clear competitive edge as the world moves towards responsible use of AI.

At Legally Wired, we offer insights that look beyond the basics of the law. Each article concludes with actionable insights to help you assess potential risks and practical ways to reduce them, empowering you to make informed, proactive choices for your business

What are RMFs?

AI Risk Management Framework is a voluntary framework from the National Institute of Standards and Technology (NIST), part of the Department of Commerce to incorporate trustworthiness considerations into the design, development, use and evaluation of AI products, services and systems.

Although voluntary, the NIST RMF guidance has been published pursuant to President Biden’s Executive Order (EO) 14110 on Safe, Secure, and Trustworthy Artificial Intelligence. It has received widespread adoption in the industry and amongst government agencies and is expected to influence the AI regulations. The NIST RMF

is to be read with NIST RMF Profile for GenAI for the specific use case. 

Businesses rely on the guidance provided by NIST RMF to 1) ensure they internally adopt a robust AI governance framework for their use of AI 2) assess vendors that may have AI components in services to ensure they align with the AI Trustworthiness standards and goals they have set for the business.

AI Governance for Marketing and Ad Agencies

Unless you happen to be an agency that has failed to harness AI capabilities in any of tour services, read on.

Use of GenAI by marketing and Ad Agencies most likely fall under one of the following use cases:

·  Use of GenAI tools to varying degrees for generation of creatives (images, videos, audio, designs)

·  Use of GenAI tools to varying degrees for copywriting

·  Use of AI for management of client accounts, services. For instance, using AI for monitoring ad accounts, monitoring performance.

(This article specifically deals with GenAI use cases)

The risks of using GenAI in your services that the NIST has identified and as a result the client is likely to assess you on are:

·  intellectual property issues, such as unauthorized use of copyrighted, trademarked, or licensed content;

·  Risk of ease in producing and access to obscene, degrading, and/or abusive content;

·   value chain and component integration challenges, such as difficulty in vetting suppliers of your vendors;

·  harmful bias in outputs that results in discrimination especially in use of GenAI in copywriting; and

·  human-AI configurations that can result in automation bias

(to note- NIST has identified a total of 12 risks associated with use of GenAI, for the purpose of this article we have culled out risks relevant to Marketing and Ad Agencies)

Hence your governance framework should be designed to mitigate the risks that are relevant to your use of GenAI. NIST guidance on how to build and integrate Trustworthiness is organized into four categories (Govern, Map, Measure and Manage) with suggested actions and the AI Actors (stakeholders in the system) that are responsible for the actions.

How to design your AI Governance Framework

The NIST RMF is a detailed guide on how to go about the exercise of adopting an AI Governance Framework that fits the use case and needs of your organization.

A key learning from our experience is that, businesses would save time and efforts in recognizing four aspects about AI governance early on: 

•  There is no one case fits all frameworks for AI governance

•  Setting up and managing a robust AI governance requires participation from various AI Actors which would be basically anyone in your organization that has access to any AI tools

• AI governance should be compartmentalized with priority frameworks (client facing functions) tackled first.