MIT’s AI Risk Repository assists business in recognizing potential AI impacts

Artificial Intelligence (AI) is transforming industries, offering unprecedented opportunities for innovation and efficiency. With these advancements, companies need to focus and understand the risks that come with AI when applied to their business. Recently, MIT teamed up with global organizations to develop an AI Risk Repository, a dynamic database containing 777 risks from 43 taxonomies.  

The primary concern highlighted in the AI Risk Repository is the lack of a shared understanding of the types of risk a business could be opening itself up to when adopting AI. 

Because there is such a large gap in knowledge of the technology and its applications, it is hard for business owners and board members to discuss and create effective policies. The repository aims to bridge this gap by creating a common frame of reference.  

The repository organizes AI risks into seven categories, each with distinct subcategories. Below are some of the most relevant risks for small-to-medium-sized businesses: 

1. Privacy & Security: AI systems can compromise privacy by leaking sensitive information or inferring confidential data without consent. This can lead to unauthorized sharing of data, identity theft, or loss of confidential intellectual property. According to the report accompanying the repository, “A notable example is the case where Samsung employees accidentally leaked confidential intellectual property to OpenAI after using ChatGPT to help with coding tasks.” 
2. Misinformation: AI systems can inadvertently generate or spread false or misleading information. The report states this could be due to “outdated knowledge bases,” “noisy training data” ,”and find tuning processes that encourage sycophantic behavior.” It is best to always double-check and verify. 
3. Malicious Actors & Misuse: AI can be used for disinformation campaigns, cyberattacks, and fraud. For example, AI-generated phishing schemes can create highly sophisticated and convincing fraudulent communications, including voice-emulation. It is always best to verify with an additional party, especially when transferring or dealing with money.  
4. Human-Computer Interaction: Overreliance on AI systems can lead to unsafe use and loss of human agency. According to the report, “If AIs take over tasks that typically require human creativity and analytical thinking, humans may engage less frequently in these cognitive processes. Over time, this may lead to a decrease in our ability to think critically and solve problems independently.” 

Understanding and managing AI risks is crucial for small-to-medium sized businesses. By leveraging resources like the AI Risk Repository, businesses can stay informed about potential risks and take proactive steps to mitigate them. As the repository states, "This creates a foundation for a more coordinated, coherent, and complete approach to defining, auditing, and managing the risks posed by AI systems"