Navigating AI’s Opportunities + Mitigating Risks
Everywhere we look now, it seems AI (artificial intelligence) is being used and talked about. The use of AI is quickly being implemented into day-to-day business. Once hailed as the key to technical innovations only seen in sci-fi novels and films, AI is quickly becoming a cost of entry into the competition for market advantage. Many companies know that they must invest in AI or risk extinction. While they aim to capitalize on AI’s opportunity, many don’t track the risks of using it.
“AI has captured our imaginations. It is expansively discussed but often misunderstood. Before leaders explore its potential, they must first understand how it works.”
– Mike Thoma, VP and Technology and Life Sciences Industry Lead at Travelers
What is AI?
AI is a broad set of digital engineering techniques that help computer programs learn. They approximate the decision-making and problem-solving that would come “naturally” to humans. These programs are meant to improve processes, speed up tasks, reduce human effort, and improve outcomes.
AI can accomplish repetitive, low-skill tasks and complex work, like reading thousands of documents in minutes while extracting information faster with fewer errors than humans can. Some examples are digital assistants that answer spoken questions (ahem, Alexa), automatic speech-to-text transcription, Chatbots for customer service, facial detection in photo software, auditing software that detects credit card fraud, medical imaging analysis, and newsfeed and product recommendations.
AI Risks
AI can be a powerful tool, but it can also be unpredictable and cause unintended consequences if not used with caution. As with any powerful technology, there are potential negative impacts and unintended consequences that must be carefully considered. Risk identification is an essential early step of any successful AI project.
The National Institute of Standards and Technology’s (NIST) AI Risk Management Framework identifies these categories of risks associated with AI systems:
Accountability Risks: AI systems making decisions or recommendations that could be influenced by an organization’s financial or business interests.
Safety Risks: AI systems producing unintended results, leading to property damage, injury or death.
Reliability Risks: System malfunction or failure to meet performance requirements.
Bias Risks: Unintentional bias in decision-making due to lack of diversity in data or incorrect data labeling.
Security Risks: Unauthorized access, modification or destruction of AI systems and their data.
Privacy Risks: Unauthorized access to private data or use of data for unintended purposes.
Explainability Risks: Uncertainty in the decisions made by AI systems and lack of understanding of the AI system’s decision process.
It’s important for companies to understand these risks and develop strategies to mitigate them. Reducing these risks will involve – you guessed it – people! Proper design and programming, data quality testing, human oversight, transparency and accountability, and continual testing and evaluation must be present to evolve the tools.
“Only with their eyes wide open can leaders launch initiatives that best take advantage of AI.”
– Mike Thoma, VP and Technology and Life Sciences Industry Lead at Travelers
Our team is here to help in assessing your overall risk profile and setting you up with the right coverage to protect yourself, your data and your assets.Contact our team at: (636) 537-5611 or send us a note at contactus@concannonagency.com today to set up your consultation to discuss more.
Source: Travelers Insurance White Paper | Pursuing AI’s Opportunities While Mitigating Its Risks
