Regaining Control of Your Data: A Battle Users Are Losing

Artificial intelligence is built on data, much of it generated by users in their everyday digital interactions. While these systems continue to advance, visibility and control over how that data is used remain limited. The balance between innovation and ownership is increasingly shifting away from the user.

Artificial intelligence has become part of our daily lives. It is no longer optional. Whether you are drafting an email or automating repetitive tasks, you are interacting with AI in one way or another, even if you are not fully aware of it.

Data Used in AI Training: Where Is the Problem?

Most of what you publish online has likely been used to train generative AI models, whether you consented to it or not. Tech companies have leveraged publicly available data under the justification of improving these systems. This raises a key question: what data is being used, and was it used legally?

Despite ongoing lawsuits and investigations, there has been little change in how companies handle this data. Users still lack full control. Your data may already be used. The real question is how to protect what remains.

Challenges in Data Control

Most AI systems operate with default opt-in behavior, meaning user input may be used for training. Opt-out options, when available, are often hidden or difficult to access.

Not all models behave the same. For example, Claude does not automatically use user inputs for training. However, once data is already used, removing it becomes extremely difficult due to technical and legal limitations.

Can You Tell If Your Data Was Used?

AI training mainly relies on text and images. Tools like “Have I Been Trained” can help identify whether images were used, but no equivalent exists for text. This highlights a major transparency gap.

Limiting Data Use on Platforms

While full control is not possible, you can still reduce exposure depending on the platform:

• Some platforms like WhatsApp, TikTok, and Reddit do not offer opt-out options.
• Others, like Facebook, Instagram, X, and LinkedIn, allow limited control through privacy settings.

Reality vs. Claims

• Despite public commitments, many companies still use data by default.
• Opt-out features are often delayed, limited, or unclear.
• Even after deletion, data may still be retained for extended periods.

Conclusion

While some control exists, it is incomplete. Transparency remains limited, and user control is still evolving.