Interested in more of our content? Sign up for our newsletter. Sign up here:
Generative AI exploded in 2023 and became an everyday technology, similar to the early stages of email and search engines. Expectations rapidly grew to the point that generative AI may be near the “Peak of Inflated Expectations” on the Gartner, Inc. Hype Cycle.
For example, generative AI private investment increased from $2.9B in 2022 to $25.2B in 2023, Fortune 500 earnings calls mentioning AI went from 266 in 2022 to 394 in 2023 and the number of foundation models (big multipurpose models that generative AI tools run on) doubled in the US, going from 51 in 2022 to 109 in 2023.
Challenges to successfully gaining the benefits from generative AI are significant despite all the investment and focus. Identifying and understanding new emerging risks can always be a challenge.
A recent HBR article suggests that AI related risks could be categorized into 4 groups based on “intent” and “usage” to aid in identifying and mitigating AI risks:
Misuse
Misapply
Misrepresentation
Misadventure
These categories are useful for some AI related risks such as:
Consumer mistrust
Training data privacy and security
Deep fakes and misinformation
Hallucinations or false informations
Inherent biases
Data usage and rights
Ethical challenges
Cybersecurity vulnerabilities and uses
Other AI related risks may fall outside of this model and therefore require more a more holistic approach such as the recent ISO standard 23894 Information technology — Artificial intelligence — Guidance on risk management. Additional risks to consider include:
Unrealistic expectations
Intellectual property infringement
Cost/ benefit to deploy and run AI models
Shadow AI initiatives
GPU shortages
Regulatory uncertainty
With the rapid rise in investment and implementation of AI in core business operations we recommend understanding the risk landscape prior to leveraging developing technologies. Risk managers and business leaders can start navigating the issues surrounding generative AI by first informing themselves and their teams. A comprehensive Enterprise Risk Management program should be built in a way that helps leaders identify, understand, manage and where necessary integrate safeguards into business strategies and operations.
Access our LinkedIn