GPT under threat by European regulators //

Italian regulators placed ChatGPT under immediate ban. OpenAI, its creator, was given 20 days to address concerns regarding the data collection and processing. The fine could be as high as $21.7 million or 4% of annual revenue, depending on which is higher.

There are indications that other European regulators could soon follow their lead. According to reports, France is now conducting its own investigation. Ireland has asked Italy for further details on the ban’s basis. The German data commissioner stated that Germany could take the same action “in principle”.

Why do we care? It was easy to forget the warnings from the legal profession that ChatGPT could be in breach of European data regulations. These regulations have, in many ways become a global de facto standard.

It could take some time for questions to be answered by the European legal system. It is clear, however, that European regulators can swiftly take action.

HTML3_ HTML4_ HTML5_ HTML6_ HTML7_ HTML8_ HTML9_ HTML10_ HTML3_

Data processing bases These are the bases:

• Consent.
• Performance of a contract
• An interest legitimate.
• Vital interest is a matter of life or death.
• It is a legal requirement.
• Public interest.

It is not clear that large-language models are being trained using data that was obtained without consent. Unless, perhaps, one assumes that AI solutions are in the public’s best interests.

Data erasure. ChatGPT’s ability to provide the “right to forget” is another challenge. In certain cases, individuals can request that their data be erased under GDPR. ChatGPT isn’t scraping the internet or collecting excessive amounts of personal data. It is however being trained on large amounts of text. OpenAI may have to ask whether it can identify the contents of those texts in terms of personal identifying information, or data it might be asked for to erase.

The post ChatGPT under Threat from European Regulators was first published on MarTech.