On May 21, Texas Attorney General, Ken Paxton, filed a lawsuit in Harrison County Court against Meta and WhatsApp data usage conduct, alleging the giant misled users about end-to-end encryption, while retaining access to private messages. Meta denies wrongdoing, as the case seeks penalties and tighter privacy controls.
Andy Stone, a Meta spokesperson, rejected the lawsuit claims, claiming the company cannot read encrypted messages. Brought under state consumer protection laws, the case seeks an injunction to stop accessing messages of Texans without consent and demands monetary penalties.
Texas has previously dealt with major tech privacy cases, including a $1.375 billion settlement with Google also over data practices.
Meta Privacy Concerns Contradict Across AI, Messaging
Recent moves show privacy issues with WhatsAppsome areas while taking it back in others.
The company paused contracts with Mercor, a $10 billion AI data startup, after a cyberattack exposed a supply chain vulnerability tied to the LiteLLM open-source library used widely across AI systems. Mercor supplies training data to Meta, OpenAI, and Anthropic, raising broader concerns about how data flows through AI pipelines.
At the same time, Meta has taken awkward steps across its own apps, which gives Meta’s moves as being hypocrite.
“Meta removed support for end-to-end encrypted chats from Instagram as of May 8, 2026,” while promoting new AI tools on WhatsApp, including “Meta adds fully private AI chats to WhatsApp.”
The WhatsApp data mining company’s Incognito Chat mode is described as “Truly private — no one can read your conversation, not even us.”
These chats disappear by default and run in isolated WhatsApp data mining systems, but critics say the language contrasts sharply with the removal of stronger protections elsewhere.
Meta also introduced Side Chat, allowing AI assistance inside WhatsApp data usage conversations using its Private Processing infrastructure.
Does WhatsApp Use Data to Train Its AI Model?
Time over time, Meta said that end-to-end encrypted WhatsApp messages between users are not used to train its AI models. But its wider AI ecosystem relies heavily on publicly available data, which goes to question WhatsApp privacy terms and Meta’s very thin line – basically invisible line – between private communication and training material.
In the European Union (EU), strict General Data Protection Regulations (GDPR) mean companies, such as Meta and Google, must offer clearer consent mechanisms and opt-out options for data scraping for AI training.
That means users in the EU and UK can formally object to their data being used in certain contexts, including training WhatsApp data mining for AI models, something not consistently available across many states in the US.
The Texas WhatsApp data usage lawsuit could encourage other states to adopt similar consumer protection actions, especially as regulators increasingly focus on how AI systems are trained and how encrypted messaging is represented to users.
Most US state level privacy laws, including California Consumer Privacy Act (CCPA) and Virginia Consumer Privacy Act (VCDPA), mainly focus on protecting personally identifiable information or restricting companies from selling user data to third parties, keeping WhatsApp privacy terms mended by what fits the company most.
Meta privacy settings abide by these rules by drawing up its own clear line between private data and public content that people post on its platforms. It argues that public posts, photos, and comments are not covered in the same way as private communications, allowing them to be used for AI training.
Because the WhatsApp privacy concerns are around data mining and usage for AI training, Meta says it is not selling data, but it keeps it internal, many state laws on data selling are not triggered.
Under European General Data Protection Regulation (GDPR) rules, Meta has been forced to implement explicit opt-out mechanisms for AI WhatsApp data usage, giving users stronger control over their personal information. Such protections are not always available in many US states, creating a regulatory gap between American and European users.
The outcome is a system that allows Meta AI data privacy policy to continue to depend on large quantities of publicly available user data while technically complying with various regional laws.
Enforcement varies dramatically by jurisdiction, with European regulators able to impose stronger obligations, and many US states adopting narrower definitions of privacy harm.
Some users that have sought to opt out of AI training data use outside of the EU have received automated responses to the effect that the privacy issues with WhatsApprequest cannot be processed in their jurisdiction, showing the uneven implementation of privacy rights.
The lawsuit also adds to a larger problem Meta has faced for years over user privacy and trust. WhatsApp terms and privacy policy do not change in the US when a concern flares up, such as the Texas lawsuit, but rather what changes is the intention behind it.
Many people have become highly concerned about how much personal information the company collects through Facebook, Instagram, and WhatsApp data usage, especially as Meta pushes deeper into AI.
For some users, the Texas case awakens fears that their online conversations and activity may not be as private as they are told.
Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Tech sections to stay informed and up-to-date with our daily articles.
