- In The Loop AI
- Posts
- ⚡Italian watchdog says OpenAI breaks GDPR
⚡Italian watchdog says OpenAI breaks GDPR
ALSO: China approves 40 AI models
Algot Persson
January 29, 2024
Happy Monday!
In Today's Article:
US wants cloud services to disclose foreign clients
China approves 40 AI models for use
Italian watchdog says OpenAI breaks GDPR
Read time: 3 minutes
Italian Watchdog: OpenAI Breaks GDPR
OpenAI
After months of investigation, the Italian authority has decided OpenAI has breached multiple articles of the General Data Protection Regulation (GDPR).
In 2023, OpenAI faced similar accusations but was able to resume its service in Italy after updating its policy.
OpenAI is facing the challenge of showing how their Large Language Models (LLMs) were trained, including ChatGPT. The Italian watchdog’s concern lies in how the data used to train the LLM was sourced.
Since it was scraped from the internet there is a question about if it was legal, since there was no explicit compliance.
If shown that OpenAI broke the GDPR, they could be fined up to €20 Million. This would also open up questions about the ethics of similar products.
🚀 In Today’s News
The US government is proposing new regulations requiring cloud service providers like Amazon, Microsoft, and Alphabet Inc.’s Google to disclose and monitor foreign clients using their platforms for AI development, particularly targeting Chinese firms. The rules would force these companies to reveal customer identities and IP addresses, similar to strict financial industry standards. This move is part of the Biden administration's broader strategy to address national security concerns and limit China's AI capabilities, potentially affecting its military advancements. The Commerce Department is seeking feedback on the proposed rule until April 29, amidst ongoing US efforts to curb China's technological progress through semiconductor export restrictions and sanctions.
China has accelerated its AI development, approving over 40 artificial intelligence models in six months. Among these, 14 large language models (LLMs) were recently authorized, including offerings from Xiaomi Corp, 4Paradigm, and 01.AI. This marks the fourth batch of approvals since Beijing mandated regulatory clearance for public LLMs in August, following initial approvals for companies like Baidu, Alibaba, and ByteDance. Chinese firms have intensified AI efforts after the global impact of OpenAI's ChatGPT. China, housing 130 LLMs, closely trails the U.S. in global AI presence. Baidu's Ernie Bot, a prominent ChatGPT competitor, has already attracted over 100 million users.
The Best AI Tools To Use/GPT’S
A/b testing AI - The easiest, fastest, and most efficient way to A/B test
AI Suggests - Get unlimited content, unlimited ideas
Julius AI AI data analyst & more
Leap AI - Build next-generation AI apps in seconds from templates
Knibble AI Get instant answers from your data using AI-powered knowledge bases
Cody - Instant Answers To Your Business Questions
guidde - Create how-to guides in seconds
HeyGen - Effortlessly produce studio-quality videos with AI-generated avatars and voices
GPTS
Become an AI & ChatGPT genius for free (Holiday Sale) - Join a 3-hour ChatGPT & AI workshop (worth $99) at a 100% discount!
Register Here: (FREE for the First 100 people)
Reply