Worried About DeepSeek? Gemini Tops the List for Data Collection Concerns

In the heated debate about the safety and privacy of artificial intelligence, especially in light of increasing scrutiny of Chinese AI models like DeepSeek, new revelations have shifted the focus. While DeepSeek certainly ignited concerns upon its launch, recent research has uncovered that some popular AI models, especially those from American companies, might be even more invasive when it comes to tracking personal data. The findings offer a surprising twist on the ongoing conversation about privacy and data security in the world of AI.

DeepSeek vs. Gemini: Privacy and Data Collection in AI Chatbots

When DeepSeek, the Chinese AI chatbot, made its debut in January, it caused quite a stir, especially in the United States. With its open-source model gaining global attention, including over 12 million downloads in just two days, fears surrounding its potential to collect and misuse personal data spread quickly. As a result, various organizations, both private and government, implemented bans against DeepSeek, driven by concerns over privacy and national security.

However, new research from Surfshark, a well-known VPN provider, has turned the spotlight elsewhere. According to their findings, Google’s Gemini chatbot is far more aggressive in collecting personal data than DeepSeek. In fact, Gemini has emerged as the largest data offender in the AI chatbot arena. While DeepSeek is often seen as a privacy risk due to its ties to China, Gemini, developed by a US-based tech giant, is responsible for collecting significantly more personal data, raising important questions about privacy practices, even in the West.

How Surfshark’s Research Unveils the Privacy Practices of Popular AI Chatbots

Surfshark’s research scrutinized the privacy practices of the ten most popular AI chatbots available on the Apple App Store. This list included familiar names like ChatGPT, Google Gemini, Copilot, Perplexity, Grok, Jasper, Poe, Claude, Pi, and DeepSeek. The study focused on several key aspects: the types of data each chatbot collects, whether it links data to individual users, and if third-party advertisers are involved in data sharing.

The results were striking. While DeepSeek was far from the worst, ranking fifth out of the ten chatbots analyzed, Google’s Gemini stood out as the most aggressive in data collection. Gemini was found to collect 22 out of 35 possible types of user data. This includes highly sensitive information such as location data, user content, browsing history, and even access to the device’s contacts list.

In comparison, DeepSeek collects an average of 11 data types, primarily focusing on user content, contacts, and diagnostics. It’s far less invasive than Gemini but still collects more data than some other competitors like ChatGPT, which gathers 10 unique data types, including usage data and contact information. Interestingly, both ChatGPT and DeepSeek allow users to manage and delete their chat history, though DeepSeek stores its data in servers located outside its home country, specifically in China.

What Undercode Says: Privacy and National Security Concerns

The concerns about DeepSeek’s data collection practices are undeniable, but the research brings to light a broader issue: the scale of data collection by AI chatbots, regardless of their country of origin. Google Gemini, the most data-intensive chatbot in the study, stands as a symbol of the ever-growing risks associated with AI development, especially when personal data is involved.

The debate over AI privacy has largely centered around geopolitical tensions, particularly between the US and China. Many have argued that Chinese tech companies, including DeepSeek, pose a unique risk due to the country’s surveillance practices and its government’s access to user data. However, as Surfshark’s findings reveal, American companies like Google are not necessarily more transparent or secure when it comes to protecting user privacy. In fact, some US-based models may collect even more data than their Chinese counterparts.

This discrepancy raises important questions: Should privacy concerns about AI be more focused on the companies that collect the most data, regardless of their location? Should users be more cautious about the data practices of major tech giants, even if they are based in countries with stronger privacy laws? As AI technology continues to advance, it is essential to weigh the trade-off between convenience and privacy, especially when chatbots and other AI tools become more integrated into daily life.

The ethical implications are vast. As AI models gather increasingly sensitive data, the potential for misuse grows. This could involve anything from targeted advertising and surveillance to data breaches and identity theft. Furthermore, the widespread use of third-party advertisers to access this data only amplifies these risks, especially when it comes to the ability to track and monetize user behaviors.

Fact Checker Results

Google Gemini collects 22 data points, far surpassing DeepSeek, which gathers 11.
Many popular AI models, including Gemini and Copilot, share sensitive data with third parties like data brokers.

3. While

References:

Reported By: https://www.zdnet.com/article/worried-about-deepseek-turns-out-gemini-is-the-biggest-data-offender/
Extra Source Hub:
https://www.instagram.com
Wikipedia
Undercode AI