Which AI chatbot is the best at protecting your privacy?
Mistral AI's Le Chat is the least privacy-invasive generative artificial intelligence model when it comes to data privacy, a new analysis has found.
Incogni, a personal information removal service, used a set of 11 criteria to assess the various privacy risks with large language models (LLMs), including OpenAI's ChatGPT, Meta AI, Google's Gemini, Microsoft's Copilot, xAI's Grok, Anthropic's Claude, Inflection AI's Pi AI and China-based DeepSeek.
Each platform was then scored from zero, being the most privacy-friendly to one, being the least-friendly on that list of criteria. The research aimed to identify how the models are trained, their transparency, and how data is collected and shared.
Among the criteria, the study looked at the data set used by the models, whether user-generated prompts could be used for training and what data, if any, could be shared with third parties.
What sets Mistral AI apart?
The analysis showed that French company Mistral AI's so-called Le Chat model is the least privacy-invasive platform because it collects 'limited' personal data and does well on AI-specific privacy concerns.
Le Chat is also one of the few AI assistant chatbots in the study that would only provide user-generated prompts to its service providers, along with Pi AI.
OpenAI's ChatGPT comes second in the overall ranking because the company has a 'clear' privacy policy that explains to users exactly where their data is going. However, the researchers noted some concerns about how the models are trained and how user data 'interacts with the platform's offerings'.
xAI, the company run by billionaire Elon Musk that operates Grok, came in third place because of transparency concerns and the amount of data collected.
Meanwhile, Anthropic's Claude model performed similarly to xAI but had more concerns about how models interact with user data, the study said.
At the bottom of the ranking is Meta AI, which was the most privacy invasive, followed by Gemini and Copilot.
Many of the companies at the bottom of the ranking don't seem to let users opt out of having prompts that they generated used to further train their models, the analysis said.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
France 24
4 hours ago
- France 24
Space race: Is Eutelsat Europe's answer to Musk's Starlink?
Two days after Russia began its invasion of Ukraine in February 2022, the amount of damage already done to internet and communication infrastructure left Ukrainian leaders with one option: tweeting a direct plea to SpaceX owner Elon Musk to activate his satellite internet service, Starlink. Starlink is the world's largest low Earth orbit (LEO) satellite constellation, providing internet to around 130 countries with over 6 million users. LEO satellites operate just a few hundred kilometres above the Earth – at lower altitudes than geostationary satellites – and play a crucial role in telecommunications, space exploration and defence as they can transmit data extremely efficiently and provide high-speed internet even in remote locations. In Ukraine, Starlink has become a lifeline, enabling civilians, military and the government to stay connected, despite ongoing attacks by Russia. 'Starlink technologies have changed this war," said Ukrainian Minister of Digital Transformation Mykhaïlo Fedorov, in November 2022. But use of the service comes with caveats, largely defined by Musk, who has threatened to withdraw access in Ukraine multiple times. In 2022, Musk denied requests from Ukraine to extend coverage into Crimea ahead of a planned drone attack on the Russian fleet, reportedly saying Ukraine was 'going too far' by seeking to use Starlink satellites for the mission. 'A matter of sovereignty' For some European leaders, this dynamic highlights the risks of relying on US tech to ensure European security, which have only increased since the January inauguration of isolationist President Donald Trump. Faced with a pressing necessity to develop home-grown alternatives, France on Friday announced a $1.55 billion investment in European telecommunications company, Eutelsat, making the French state the company's largest shareholder. "It's a matter of sovereignty," French President Emmanuel Macron said as he warned that Europe was on the verge of being 'completely' squeezed out of the LEO satellite constellation market. "SpaceX has disrupted the market, Amazon is also getting involved. China is not far behind, and I think we all need to be very clear-headed," Macron said. "Eutelsat is the only non-American and non-Chinese player with these low-earth orbit constellations and such a wide range of services.' Britain also holds a 10.9 percent share in Eutelsat, and Germany now pays for Ukraine to access the service. In April 2025, Berlin set up 1,000 Eutelsat terminals in Ukraine with the goal of providing an alternative, rather than a replacement for the 50,000 Starlink equivalents in the country. 'When you're in a situation as you see in Ukraine today, you want to have all your options open,' Eutelsat's chief executive, Eva Berneke, told Reuters. Playing catch-up With around 650 satellites in orbit, Eutelsat is the only other largescale LEO satellite provider in the world, but still lags far behind Starlink's 7,500 units. Closing the gap between the two companies is more than a matter of financial investment. 'Although Europe is trying to catch up, it's a way behind in terms of the technology,' says Barry Evans, professor of satellite communication at the University of Surrey, UK. In the US, Starlink is already providing a service that beams internet direct to handheld devices – a capability that does not yet exist in Europe. While Eutelsat provides a more reliable service as its terminals must be professionally installed, Starlink's satellites are located at lower altitudes, thereby reducing latency. And as much as European leaders see Musk as a liability, Starlink's biggest advantage may be its owner. 'Starlink is part of the Musk empire and very much involved with SpaceX,' Evans says. 'Musk produces the satellites. He produces the launchers. He produces the terminals. The whole system is vertically integrated.' Synergy between SpaceX and Starlink means innovations such as the partially reusable Falcon launcher can be used to rapidly scale up the number of satellites Starlink sends in to space. The company aims to launch some 42,000 satellites into its LEO constellation overall. Transforming Eutelsat into a direct competitor is, currently, a tall ask. Evans says, 'it's difficult to see how Europe can really catch up on that'.
Local France
8 hours ago
- Local France
France's Versailles unveils AI-powered talking statues
Versailles has announced a partnership with US-based OpenAI and French start-up Ask Mona to bring a 21st-century AI touch to the iconic 17th-century palace. Curious visitors can delve into Versailles' rich history by scanning a QR code next to one of some 20 garden statues – triggering interactive conversations in French, English, or Spanish. 'The Palace of Versailles is now testing artificial intelligence, whose tremendous capabilities will greatly enrich the visitor experience,' said the museum's president, Christophe Leribault. OpenAI and Ask Mona said it was a 'golden opportunity' to showcase their technology in a world-famous location that welcomes some eight million visitors annually. 'Whether you're a heritage expert, a museum curator, or a visitor setting foot in the gardens of the Palace of Versailles for the first time, there's something for everyone,' said Julie Lavet, OpenAI's head of European partnerships. Advertisement And Ask Mona's president said this partnership is a chance to highlight AI's lesser explored applications. 'Often, when we think of artificial intelligence, we think of it in terms of productivity, but here, it's really artificial intelligence that is a lever for curiosity,' said Marion Carre.
Euronews
13 hours ago
- Euronews
Which AI chatbot is the best at protecting your privacy?
Mistral AI's Le Chat is the least privacy-invasive generative artificial intelligence model when it comes to data privacy, a new analysis has found. Incogni, a personal information removal service, used a set of 11 criteria to assess the various privacy risks with large language models (LLMs), including OpenAI's ChatGPT, Meta AI, Google's Gemini, Microsoft's Copilot, xAI's Grok, Anthropic's Claude, Inflection AI's Pi AI and China-based DeepSeek. Each platform was then scored from zero, being the most privacy-friendly to one, being the least-friendly on that list of criteria. The research aimed to identify how the models are trained, their transparency, and how data is collected and shared. Among the criteria, the study looked at the data set used by the models, whether user-generated prompts could be used for training and what data, if any, could be shared with third parties. What sets Mistral AI apart? The analysis showed that French company Mistral AI's so-called Le Chat model is the least privacy-invasive platform because it collects 'limited' personal data and does well on AI-specific privacy concerns. Le Chat is also one of the few AI assistant chatbots in the study that would only provide user-generated prompts to its service providers, along with Pi AI. OpenAI's ChatGPT comes second in the overall ranking because the company has a 'clear' privacy policy that explains to users exactly where their data is going. However, the researchers noted some concerns about how the models are trained and how user data 'interacts with the platform's offerings'. xAI, the company run by billionaire Elon Musk that operates Grok, came in third place because of transparency concerns and the amount of data collected. Meanwhile, Anthropic's Claude model performed similarly to xAI but had more concerns about how models interact with user data, the study said. At the bottom of the ranking is Meta AI, which was the most privacy invasive, followed by Gemini and Copilot. Many of the companies at the bottom of the ranking don't seem to let users opt out of having prompts that they generated used to further train their models, the analysis said.
