A nonprofit artificial intelligence ethics organization has asked the US Federal Trade Commission to investigate Open AI and regulate some of its activities.

At the heart of CAIDP’s (the Center for AI and Digital Policy) complaint is the GPT-4 model that OpenAI released earlier this month. OpenAI has made GPT-4 available to its ChatGPT Plus customers. Developers access the model through an API.

GPT-4 is the most advanced neutral network OpenAI currently has. It can generate text and code, as well as perform other complex tasks such as: B. detecting hacking attempts. According to OpenAI, GPT-4 passes a mock bar exam with a score of about the top 10 percent of test takers. The previous GPT-3.5 model scored in the bottom 10 percent of testers.

Complaints about unfair and deceptive business practices

According to the complaint, the commercial release of GPT-4 violates Section 5 of the FTC statute. This part of the law prohibits unfair and deceptive business practices. CAIDP claims that OpenAI is trying to shirk responsibility for the unfair and deceptive use of its technology.

“The company seeks to object to unlawful, deceptive, unfair and dangerous uses of its product that would be apparent to many users,” the CAIDP complaint reads. “It would be irresponsible for any company in any other market sector to sell a product to the public that has so many known risks and then try to deny responsibility.”

To support its argument, CAIDP in the complaint cites a recent statement by the FTC that “merely warning customers of abuse or asking them to provide information is hardly enough to deter adversaries.”

research and regulation

In its complaint, the CAIDP calls on the FTC to open an investigation into OpenAI. In addition, the company wants to suspend “future commercial use of GTP” until OpenAI meets the FTC’s guidelines for AI products.

In addition, CAIDP proposes a number of regulatory moves, including mandating an “independent review throughout the lifecycle of GPT AI” and creating a mechanism to report rule violations related to GPT-4. It also calls on the FTC to establish fundamental standards for the generative AI market.

Other organizations are affected

Although AI says it has asked outside experts to assess the potential risks of GPT-4, CAIDP is not the first group to worry about the rapid pace of developments in AI. For example, Arstechnica also reports that the Future of Life Institute has issued an open letter urging AI labs to immediately suspend training of AI systems more powerful than GPT-4 for at least 6 months. This letter was signed by a number of notable names, including Elon Musk and Steve Wozniak.