The organization alleges that the creator of ChatGPT has violated Section 5 of the Federal Trade Commission Act with the release of GPT-4. This section of the act prohibits any practices or acts considered deceptive or unfair and that affect commerce.

CAIDP has complained to the US Federal Trade Commission, intending to prevent the distribution of advanced AI systems to the general public. The complaint focuses on GPT-4, the latest large language model developed by OpenAI, which CAIDP has characterized as “partial, misleading, and posing a threat to public safety and privacy” in its filing with the FTC on 30th March.

CAIDP, a non-profit independent research organization, has contended that the introduction of GPT-4 for commercial purposes goes against Section 5 of the FTC Act, which forbids any practices or actions that are deceitful or unjust and have an impact on commerce.

To support their argument, the Artificial Intelligence ethics group referred to the information in the System Card for GPT-4, which includes the following: According to the System Card for GPT-4, the model can strengthen and replicate particular prejudices and perspectives, such as derogatory and detrimental associations with specific underrepresented communities.

The System Card for GPT-4 cautioned that AI systems could reinforce entire belief systems, perspectives, truths, and untruths and make them fixed or unalterable, which could impede future inquiry, introspection, and progress.

Risky GPT-4 Release without Assessment

CAIDP stated that OpenAI was fully aware of the potential risks of GPT-4 when releasing it for public use and that no third-party evaluation of GPT-4 was conducted before its launch. Consequently, CAIDP has requested that the FTC investigate the offerings of OpenAI and other developers of advanced AI systems.

CAIDP has called for the FTC to take action, stating that it is time for the regulatory body to step in. The organization has requested that the FTC launch an inquiry into OpenAI, prohibit any additional commercial distribution of GPT-4, and put in place adequate safeguards to safeguard consumers, companies, and the market.

Although ChatGPT-3 was made available in November, GPT-4 is the most recent version and is estimated to be more than ten times better and more capable. According to a study published on 14th March, GPT-4 can perform exceedingly well, achieving scores within the top 90th percentile in the most challenging law and high school exams in the United States.

The petition, signed by tech luminaries including Apple’s Steve Wozniak and Elon Musk, calls for a “pause” in advancing AI systems more potent than GPT-4. This complaint coincides with the petition. Marc Rotenberg, the president of CAIDP, was one of the two thousand six hundred signatories of the petition filed by the FLI (future of life institute) on 22nd March.

The petition’s authors maintained that advancing advanced Artificial Intelligence could mark a significant shift, whether for the worse or, the better, in the antiquity of existence on our planet. UNESCO has urged nations to follow the United Nation’s Recommendation on the integrity of the Artificial Intelligence framework, asking them to implement it.

A former Artificial Intelligence scientist for Google has argued that Google’s Artificial Intelligence chatbot, named “Bard,” was designed using the responses generated by ChatGPT. Although the researcher who made the allegation has resigned, Google executives have refuted the claims made by their former employees.