Group accuses OpenAI of violating trade laws

A group focused on AI ethics, addressing questions in AI, urged the U.S. Federal Trade Commission (FTC) to launch an investigation into OpenAI.

Researchers at the nonprofit Centre for AI and Digital Policy (CAIDP) say that the release of GPT-4 breaches trade laws. In their view, the product deceives consumers and exposes them to risk.

Experts cited Section 5 of the Federal Trade Commission Act. Under the document, “unfair or deceptive acts or practices in commerce, or acts or practices that affect commerce, are prohibited.”

CAIDP said that OpenAI violated the law by releasing the product. In their view, the company knew of the dangers but ignored them.

In the complaint, the centre urged officials to study and possibly regulate OpenAI’s generative artificial intelligence technology, particularly its latest model GPT-4.

“The Federal Trade Commission said that the use of AI must be transparent, explainable, fair and empirically grounded, contributing to accountability.” The OpenAI GPT-4 product does not meet any of these requirements,”

the centre said.

The authors of the complaint argued that there should be an independent oversight body and an evaluation of commercial AI products available in the United States. They also call on suspending deployments of GPT-based services until they meet the law.

“CAIDP calls on the FTC to commence an investigation into OpenAI, ban further commercial releases of GPT-4 and ensure the creation of needed barriers to protect consumers, businesses and the commercial market,” the statement said.

The group also highlighted the shortcomings and risks noted in a report describing the model’s capabilities.

OpenAI acknowledged that GPT-4 is far from perfect. It can reinforce stereotypes, generate harmful text and spread false information. Sam Altman said that he was ‘a little frightened’ by the technology’s impact on society.

Additionally, the centre criticised the company for a lack of transparency regarding the model and training data.

In March, more than a thousand AI experts urged to suspend the development of large language models for six months. Among the signatories was CAIDP director Marc Rotenberg.

Two days later, another group of industry experts criticised the initiative.

In February the FTC urged developers to market AI products responsibly. Otherwise, unscrupulous entrepreneurs warned they would be punished.