Italy becomes the first country to ban ChatGPT! OpenAI CEO: I don't think there's any violation

Just yesterday, the Italian government suddenly ordered to ban ChatGPT!

Soon, OpenAI CEO Sam Altman responded that the company had shut down the local ChatGPT service (but it believed that it had not violated any regulations).

Since then, Italy has become the first country in the world to announce the ban on ChatGPT.

(ChatGPT cannot be used in other countries and regions because OpenAI did not open it from the beginning.)

Italy bans ChatGPT and fully investigates privacy issues

On March 31, Italy’s national privacy regulator officially ordered the ban of ChatGPT and accused OpenAI of “illegal collection of personal data.”

It is worth noting that this “temporary ban” will remain in effect until OpenAI can respect the European Union’s landmark privacy law, the General Data Protection Regulation (GDPR).

File address: https://www.garanteprivacy.it/home/docweb/-/docweb-display/docweb/9870847 #english

According to a press release issued by the Italian national data protection agency GPDP, OpenAI’s behavior of collecting user data and using it for algorithm “training” lacks legal basis.

At the same time, the leakage of user conversation and payment information on March 20 also clearly exposed OpenAI’s problems with the processing of personal information.

In addition, although OpenAI claims that ChatGPT is aimed at users over 13 years old, due to the lack of corresponding age verification mechanism, minors are likely to see content beyond their age. Information about one’s own development and level of self-awareness.

The regulatory agency stated that OpenAI has 20 days. If it does not provide feasible remedial measures, it will face a fine of up to 20 million euros, or 4% of the total global annual turnover. .

In response, OpenAI stated that it has now closed the service for Italian users and promised to protect the privacy of users:

"We are When training systems like ChatGPT, we have been working hard to reduce personal data, because we want this AI to understand the world, not individuals. Of course, we also believe that AI supervision is necessary."

It seems that the future fate of ChatGPT in Italy may depend on these 20 days.

However, this is not the first time for Italy to take such action against AI chatbots.

In February, regulators banned chatbot app Replika.ai. Replika was "notorious" for the intimate relationships some users formed with robots. Later, Replika removed the option for pornographic role-playing, and some users even required help from a suicide hotline.

Should AI tools be restricted?

After the release of ChatGPT, although the ability to write papers and code is dazzling, its spread of misinformation, impact on employment, and broader risks to the entire human society have all made people dazzled. Many people are worried and issuing warnings.

Recently, there have been waves of bans on GPT technology.

Not only Turing Award winner Yoshua Bengio, Musk, Marcus and other thousands of big guys signed an open letter, calling on all AI laboratories around the world to suspend training on machines more powerful than GPT-4 AI model.

Moreover, the consumer rights protection organization BEUC has also recently called on the EU and national authorities to investigate ChatGPT.

In addition, many experts have also stated that new regulations are needed to manage AI due to its potential impact on national security, employment, and education.

A spokesman for the European Commission said: "We hope that all companies active in the EU will respect the EU's data protection rules. Enforcing the General Data Protection Regulation is the responsibility of the EU data protection authorities."

European Commission Executive Vice President Margrethe Vestager said on Twitter that the Commission, which is discussing the EU Artificial Intelligence Bill, may not be inclined to ban artificial intelligence.

"No matter which technology we use, we must protect our own rights. This is why we do not regulate AI technology, but regulate the use of AI."

In the United States, the non-profit organization Center for Artificial Intelligence and Digital Policy (CAIDP) also asked the Federal Trade Commission (FTC) to investigate whether OpenAI violated consumer protection rules and ban it It further released GPT-4.

After all, with the precedent of a joint letter from thousands of big bosses, CAIDP’s appeal seems more logical. Moreover, CAIDP Chairman Marc Rotenberg surprisingly signed the letter.

File address: https://cdn.arstechnica.net/wp-content/uploads/2023/03/CAIDP-FTC -Complaint-OpenAI-GPT-033023.pdf

Overall, the content of the CAIDP complaint is consistent with that letter, calling for slowing down the development of generative AI models. , and implement stricter government oversight.

However, CAIDP further believes that ChatGPT is "biased, deceptive, and poses a risk to public safety."

Although OpenAI has publicly pointed out the potential threats of AI-generated text, CAIDP believes that GPT-4 has gone far beyond the line and regulatory action should be taken to protect consumers.

CAIDP stated that the GPT-4 model has many major potential threats, such as the way GPT-4 can generate malicious code and highly customized propaganda, and its biased training data, This can lead to stereotypes or unfair racial and gender preferences in areas such as recruitment.

CAIDP believes that OpenAI’s actions violate Section 5 of the FTC Act, which prohibits unfair and deceptive trade practices.

Among them, the illusion of the AI ​​model and the nonsense are all deception in the view of CAIDP. It warned ChatGPT of "promoting deceptive business statements and advertising," which could fall under the FTC's purview.

In the complaint, CAIDP asked the FTC to halt all further commercial deployments of the GPT model and to require that the model be reviewed before any future rollout. Conduct an independent evaluation. It also calls for a publicly accessible reporting tool, similar to the one that allows consumers to file fraud complaints.

CAIDP looks for signs that big tech companies are trying to monopolize competition, and the investigation into OpenAI, one of the major players in the AI ​​arms race, marks a major step forward in the investigation, said CAIDP chairperson Lina Khan. upgrade.

Some analysts believe that although CAIDP’s complaint will not cause the FTC to take action in the short term, relevant supervision is likely to occur in the future, and may allow GPT-5 to Release postponed.

Finally, I can only say that the "reputational risk" that Google has been worried about has finally arrived.

Reference: ​

​​https: //www.php.cn/link/d35a29602005cb55aa57a5f683c8e0c2​​

The above is the detailed content of Italy becomes the first country to ban ChatGPT! OpenAI CEO: I don't think there's any violation. For more information, please follow other related articles on the PHP Chinese website!