Artificial intelligence (AI) began as a specialized tool requiring coding knowledge in order to use, but is now generally accessible to anyone due to its ability to interact using plain language. Many now use AI, but few understand what it is.
Generally speaking, AI is a technology that enables computers to mimic human learning, analysis and decision-making. AI does this by analyzing vast amounts of data, identifying patterns in the data and learning from those patterns. Generative AI (“GAI”) is a subset of AI that focuses on generating content, such as text or images, based on the learned patterns and then predicting the correct response to a user’s prompt. GAI takes commands from its users, which can be formed as a question or as a directive. In either case, GAI is not trained to say, “I don’t know the answer to that.” It is trained to analyze its dataset for patterns and generate a predictive response to a prompt based on those patterns.
Businesses use GAI frequently. Below is a Top 10 List of legal issues surrounding the use of AI by businesses.
- Contracts — GAI handles narrow, discrete issues much better than broad or complex ones. In contracts, GAI’s shortcomings include a lack of context (i.e., negotiating power of the parties, institutional knowledge, acceptable business terms) and by prioritizing fluency over form GAI frequently fails to comply with law (i.e., state-specific rules regarding how text must appear in order to effectively disclaim warranties or waive a right to jury trial).
- Confidentiality — Many GAI models allow a user’s interactions with GAI to form part of the GAI’s dataset. Thus, the interactions could later be used in GAI’s response to another user’s prompt. Feeding confidential or personal information into GAI should be avoided.
- Accuracy — GAI prioritizes fluency as opposed to fact-checking its responses. When GAI lacks the information in its dataset to respond accurately, it can “hallucinate” false information in order to respond fluently to the prompt. Needless to say, by relying on false information in making a business decision a business may be exposed to liability.
- Employment Discrimination — Employers use AI tools to automate and enhance HR functions, including the review of applicants, performance evaluations and assisting in matching candidates with well-suited jobs. AI may be biased in its dataset or decision-making. If relied on by an employer, this underlying bias may open the employer up to liability under anti-discrimination laws.
- Consumer Interactions — Businesses use AI chatbots to interface with customers. Chatbots may collect data without consent, provide misleading responses or interact with individuals in other ways that violate law.
- Litigation — A user’s interactions with AI may not be protected from disclosure in discovery. Businesses should be cognizant of the potential risks associated with AI interactions being used against the business in later litigation.
- IT Security — Hackers are utilizing AI technology as well. Phishing e-mails that previously had grammar or spelling errors no longer do because hackers use GAI to draft emails. GAI videos also may be used to mimic an individual’s likeness or voice in order to infiltrate a business.
- Regulations — Regulations change from administration to administration. Under President Joe Biden, the Federal Trade Commission, Equal Employment Opportunity Commission and Office of Federal Contract Compliance Programs issued guidance and took enforcement actions regarding the use of AI. President Donald Trump rescinded that guidance in favor of promoting the use of AI. As the presidency and governorship changes, AI regulations will change as well.
- Legislation — Legislation addressing AI typically identifies “high-risk” areas for use of AI in making a decision. High-risk areas include: education, employment, lending, legal services, health care, housing, insurance and government services. You must be aware of special rules regarding the use of AI if you operate in a jurisdiction with AI-legislation.
- More to Come — AI and GAI are still new to most of us. The law is still changing and reacting to AI. As the law changes, new legal issues will undoubtedly arise. Businesses must remain alert to changes and agile in response to them.
Please contact a MacDonald Illig attorney at 814/870-7600 or visit macdonaldillig.com if you have any questions regarding the use of AI in your business.