Google's sneaky AI move

Reading time: 5 minutes  

 Newsletter | News | Tools 

Today we will discuss-

• 🌨️U.S. calls cloud giants to monitor foreign AI activity

• 📱Google AI to read your private messages

• 🚨Italian DPA finds ChatGPT in violation of Europe's privacy laws

• 😎9 amazing AI tools you might not have heard of

All this and more - Let's dive in!

👩‍🍳What’s cooking in the newsroom

Biden Administration to restrict Chinese firms from using American data centers for AI training

☕News - The United States is urging cloud service providers like Amazon and Microsoft to investigate and report foreign clients developing AI on their platforms, signaling an escalation in the tech rivalry with Beijing.

The Biden administration's proposal, set to be released on 29 January, requires companies like Amazon and Google to share the names and IP addresses of their foreign customers. The draft rule, published on Sunday, also mandates these companies to allocate a budget for collecting such information and reporting any suspicious activity.

⛩️What this means for China - If put into action, these rules could allow the U.S. to restrict a significant channel through which Chinese companies obtain access to essential data centers and servers for AI training and hosting, thereby curbing AI development with potential military capabilities.

🕵🏻‍♂️What prompted this move though? China's advancements in AI are a significant worry for the U.S., which regards Beijing as its primary global strategic competitor.

In October, the U.S. tightened controls on AI chip exports to block China from getting advanced chips directly. However, the rules did not prevent Chinese firms from accessing the same chip capabilities through the cloud. To this end, Commerce Secretary Gina Raimondo stated, "These models getting in the hands of non-state actors or people that aren't our allies is very dangerous."

Google update reveals Bard will analyze private messages from your past

🤖News - Recently, Google revealed plans to integrate Bard into its primary messaging platform, Google Messages. However, a potential downside came to light with a recent update, revealing that Google's AI will now analyze and read your private messages, including ones from the past.

🤔Should users be concerned? While Bard is designed to be your personal AI assistant within Messages, offering benefits in communication, creativity, and information, users should be mindful of how this might affect their privacy.

Bard's analysis of private messages would  extend to understanding the context, tone, and sentiment to tailor responses. Furthermore, Bard would also examine the user's message history with different contacts to personalize responses according to their specific relationships.

Additionally, when users make requests to Bard in Messages, the requests will be sent to the cloud for processing, possibly viewed by humans in an anonymized form, and stored for 18 months. Even if you disable the AI, this data will remain for a few days, though manual deletion is an option.

More importantly, these processes fall outside the newly implemented default end-to-end encryption of Google Messages. What this simply means is when Bard analyzes your messages or stores data, it doesn't follow the new privacy features in Google Messages. So, when you interact with Bard, you're essentially talking directly to Google. This is similar to other AI chatbots, where your queries may not be completely private. 

Italian Data Protection Authority flags OpenAI's ChatGPT for violating Europe's privacy laws

🛡️News - The Italian Data Protection Authority has notified OpenAI that it is suspected of breaching European Union privacy regulations. This comes after a lengthy investigation into its AI chatbot, ChatGPT. 

🤓For context - Last year, the Italian authority expressed concerns about OpenAI's compliance with the EU's General Data Protection Regulation (GDPR). This led to a temporary ban on local data processing for ChatGPT, resulting in the AI chatbot being temporarily suspended in the market. 

In a March 2023 communication, including the lack of a proper legal basis for collecting and processing personal data for training ChatGPT's algorithms, the AI tool's tendency to produce inaccurate information (hallucinate), and concerns about child safety.  

After addressing some of these concerns, OpenAI swiftly resumed ChatGPT's service in Italy last year. However, the DPA continued investigating and has now reached preliminary conclusions that the tool is in violation of EU law.

While the Italian authority hasn't specified which of the previously suspected ChatGPT breaches it has confirmed, a key issue revolves around OpenAI's claimed legal basis for processing personal data to train its AI models. Simply put, ChatGPT was developed using publicly scraped data, including personal information, which violates EU laws, as a valid legal basis is required for processing individuals' data in the EU.

👩🏼‍💻What else is happening?

• Japan, U.S. agree on AI research for drones to assist new fighter jet

• China accelerates AI race, approves over 40 models for public use in six months

• AI-designed drug for inflammatory bowel disease enters human clinical trial

• Samsung to showcase Baidu’s Ernie AI in latest Galaxy phones

• Netflix says Generative AI can adversely impact its operations

 🤹🏻‍♂️Interesting facts

• Human orchestras have played music composed by AI. According to Classic FM, an online digital radio platform, musicians from around the world collaborated to perform a live orchestral score composed entirely by AI in Armenia in October 2019.

• In 2018, an artwork created by AI was sold at Christie's for $432,500, making history in the art world. This achievement surpassed pre-sale predictions and highlighted AI's significant creative potential.

• As per a Statista report, the marketing industry has the highest generative AI adoption rate, with content generation for marketing purposes being the most common use case for ChatGPT.

 😵Discover mind-blowing AI tools

1. Cleanvoice AI - A tool for audio editing, removing filler sounds, stuttering, mouth sounds, and other artifacts from audio recordings

2. Kansei.app - An AI language learning platform that offers personalized language practice through immersive conversations with lifelike personas

3. Charlie - A tool that provides a personalized and friendly way to stay informed about the news

4. Publer - An all-in-one social media management tool

5. Scite - An AI-driven research tool that assists users in finding reliable information from millions of research articles

6. ConnectedFlow - An AI-powered platform that helps eCommerce businesses optimize their promotions

7. Freed AI - A tool that listens, transcribes, and writes medical documentation for clinicians

8. Cactus interior - Allows users to upload their room pictures and redesign them with a fresh and unique style

9. FlyFin - An AI-powered tax service for freelancers and gig workers

💥Mythbuster

Myth: AI understands right and wrong by itself.

Fact: This is not true. AI models operate based on patterns and data they were trained on and don't have a built-in moral compass. Thus, human developers must explicitly teach ethical guidelines during creation and use to ensure responsible AI behavior.