Privacy Options for AI: Protection for Consumers is Lacking
On September 22, 2026, OpenAI, Anthropic, and Google introduced privacy options for their chatbots, but these options are not available to regular users. Cryptographer Moxie Marlinspike draws attention to the new AI service Confer, which ensures better privacy through technical measures. This topic is of great importance, as chatbots often store sensitive conversations that can be requested by companies or law enforcement through legal requests. Marlinspike emphasizes that privacy issues have shifted from messaging apps to AI interactions.
The most well-known chatbots, such as ChatGPT, Claude, and Gemini, offer virtually no real privacy. OpenAI, Anthropic, and Google provide the ZDR contract for enterprise clients, allowing chat logs to be deleted immediately after processing. However, this option is not available to regular consumers, and there are exceptions. For instance, Anthropic states that they do not provide ZDR support for certain models due to the risk of abuse. This raises the question of how safe consumers truly are when using these AI services.
Marlinspike points to the risk that people intertwine their personal lives with AI applications, which can lead to significant privacy concerns. He highlights the need for technical guarantees rather than just policy interests. While services like Proton and DuckDuckGo promise that conversations remain private, robust end-to-end encryption is lacking. However, services like Confer utilize technical solutions such as a trusted execution environment (TEE) for better data security. This distinguishes Confer from other providers that lean more on policy than on technical solutions.
The discussion about ensuring privacy in AI comes with costs. Confer employs a freemium model, where the lowest paid subscription price is significantly higher than that of other providers. Marlinspike explains that the higher costs represent the true price, without blog fees or other financial models. This raises questions about the sustainability of privacy services unless they focus entirely on data monetization or advertisements. In the current climate, it is clear that true privacy in the world of AI is not cheap.
Read the full article from AI Insider.
Gerelateerde artikelen
Anthropic accelerates AI development through self-learning systems
AI company Anthropic has recently made significant progress in the development of artificial intelligence, now achieving in one year the same amount of progress that previously took two years. This re...
Tech Festival
06 November 2026
Amazon introduces text feature for Alexa
Amazon has announced that it will soon be possible to send text messages to Alexa to get help with shopping. This feature will be rolled out to millions of users in the coming weeks. This marks an imp...
Tech Festival
06 November 2026
Return of the Glowing Logo in MacBook Pro
Apple has recently brought back the glowing logo on the new MacBook Pro, featuring the PROLED XDR display. After a decade-long absence, this reintroduction offers tech enthusiasts something to look fo...
Tech Festival
06 November 2026