Elon Musk’s satellite internet company Starlink has quietly revised its privacy policy to permit the use of customer data for training artificial intelligence models. The change, which took effect on January 15, raises fresh concerns about data privacy and surveillance as tech companies race to develop more powerful AI systems.
The updated Global Privacy Policy now states that unless customers actively opt out, Starlink may use their data to train machine learning or artificial intelligence models. The policy also allows the company to share this information with service providers and what it calls third-party collaborators, though Starlink does not specify who these partners are or how the data might be used.
What Data Does Starlink Collect
Starlink gathers extensive information from its more than 9 million users worldwide. The company collects location data, credit card details, contact information, and user IP addresses through its network of over 9,000 satellites that provide internet connectivity.
Beyond basic account information, Starlink also captures what it describes as communication data. This category includes audio and visual information, data contained in shared files, and inferences the company may draw from other personal information it collects. The privacy policy does not clarify exactly which types of data will be used for AI training purposes.
Privacy Advocates Raise Alarms
The policy revision has sparked criticism from consumer rights groups and privacy experts who warn that using personal data to train AI systems expands surveillance capabilities and creates new opportunities for misuse.
Anupam Chander, a technology law professor at Georgetown University, expressed concern about the vague nature of the policy changes. He noted that while there may be legitimate uses for customer data, the policy lacks clear limits on how the information will be applied. Chander said the update would make him worried if he were a Starlink user.
Connection to xAI Merger Talks
The privacy policy change comes as SpaceX, Starlink’s parent company, engages in discussions about a potential merger with xAI, another Musk-owned venture focused on artificial intelligence development. The talks are taking place ahead of a planned initial public offering for SpaceX later this year that could value the company at more than $1 trillion, making it even more valuable than it already is as the world’s most valuable private company.
xAI, which was most recently valued at $230 billion following a recent funding round, currently develops the Grok large language model chatbot and also owns X, the social media platform formerly known as Twitter. A merger between SpaceX and xAI would give the AI company access to vast new datasets from Starlink’s millions of users, potentially accelerating development of AI-powered services.
How the Policy Changed
An archived version of Starlink’s privacy policy from November, reviewed by news outlets, contained no language about using customer data to train AI models. The January 15 update introduced the new provisions allowing AI training and third-party data sharing.
The company did not respond to requests for comment about the policy changes or provide additional details about which third-party collaborators might receive access to customer data.
Users Can Opt Out
Starlink has built an opt-out mechanism into the new policy rather than requiring users to opt in. Customers who wish to prevent their data from being used to train AI models operated by third-party collaborators must navigate to their user settings and select the relevant option.
The privacy policy indicates that opting out specifically applies to third-party AI training but does not clarify whether it prevents Starlink itself from using customer data for its own AI development purposes. The policy states that users can email the company at privacy@spacex.com to exercise various privacy rights, including opting out of certain data uses.
Broader Implications for Tech Industry
The Starlink policy update reflects a growing trend among technology companies seeking to leverage user data for artificial intelligence development. As AI systems require massive amounts of information to improve their capabilities, companies with access to large user bases are increasingly modifying their terms of service to permit AI training.
The move also highlights tensions between rapid AI advancement and consumer privacy protections. Critics argue that opt-out policies place the burden on users to protect their own data rather than requiring companies to obtain explicit consent before using personal information for new purposes.
SpaceX already operates as the world’s most valuable privately held company, and the addition of AI capabilities could further cement its position across multiple technology sectors. The potential merger with xAI would create synergies between satellite internet infrastructure and artificial intelligence development, giving the combined entity significant advantages in data access and AI model training.
