The question many are asking is direct and pertinent: Can the police access my ChatGPT conversations? or with other attendees such as Gemini on AndroidWith millions of people using these services as if they were digital confidants to vent, ask for advice, or even snoop on sensitive topics, it's no wonder that anxiety has skyrocketed.
The debate has escalated after viral cases and announcements from the tech companies themselvesOn one hand, Spanish legal experts are reminding us what is and isn't allowed here under the law. On the other hand, OpenAI has detailed that monitors and reviews conversations in the face of serious risks And, in extreme situations, it could alert the authorities. In between, there remains the eternal balancing act between Privacy & Security.
The viral video that lit the fuse
On TikTok, creator @_mabelart shared a storytime in which she claimed that the police had summoned and questioned her. for their searches and conversations on ChatGPTShe recounted that, as a true crime enthusiast, she had asked the chatbot raw questions such as How long does it take for a body to dissolve in acid? How to remove DNA from a crime scene? or what would happen to to bury a corpse in a forestAccording to his account, he soon received a notification from the court to testify.
The story sparked all sorts of comments. Some people took it literallySome questioned it, others did, and the protagonist herself clarified that it could be part of a game with her community. Beyond the veracity of that specific case, the video served to raise the key question: Do the authorities have access to what we discuss with an AI? Like ChatGPT or Gemini?
What can and can't the police in Spain do with your chats on ChatGPT?
Attorney Jesús P. López Pelaz, founder and director of Bufete Abogado Amigo, is clear: Law enforcement agencies cannot directly "enter" your chats with ChatGPT or other language models. If you want to obtain information, you must follow specific channels and guarantees.
According to this expert, There are two possible ways to access information related to your interactions with AI, always with legal support:
- Logs on your own device (computer or mobile): they will need your consent or a court order to search, and it's good to know how limit access to certain chats.
- Logs on the provider's servers (the ISP or AI company): a court order directed to the company is required to provide that data.
It is important to emphasize that, in both cases, the typical access would be to records or traces (logs) linked to the activity, not simply reading "all your complete conversations" as if it were a personal chat. A specific permit is required.and furthermore, in practice, it is required to gather sufficient evidence of a crime to justify the measure before a judge.
When the supplier is located abroad, things get complicated: A Spanish court order is not enough.It must be communicated to the authorities of the relevant country (often the United States), who will assess whether the request is appropriate, justified, proportionate, and based on solid evidence. International cooperation and deadlines They are, therefore, a relevant factor.
Is this a “private communication” like a chat between people?
The legal distinction is significant. López Pelaz points out that Interacting with an AI is more like an internet search. than a protected conversation between two individuals. That is, it is not classified in the same way as a private communication between people for the purposes of confidentiality of communications. What you do with a chatbot It is considered an information society service: you send a request and receive an automated response.
Incognito mode and clearing history: what to do and what not to do
Another common misconception: Use incognito mode or clear your browsing history. It doesn't make your interactions invisible to a provider. That prevents browsing from being saved on your device, but It does not delete records on the company's servers.With a court order, investigators could request that information from the provider, regardless of whether you no trace is seen in your browserTo improve your protection, consult guides on Privacy & Security.
When might the authorities review your interactions?
Theoretical scenarios are limited and require a legal path. Essentially, there are two possible scenarios mentioned by experts when discussing police access:
- When there is an express judicial authorization that motivates access to that data.
- To prevent the commission of a crime at that very moment if a real and imminent risk is proven, within the applicable legal framework.
Even in that case, It does not imply an open bar or mass surveillance.These are specific, justified, and judicially controlled procedures that may also require cooperation with foreign authorities when the company that holds the records operates outside of Spain.
What does OpenAI say about reviewing conversations on ChatGPT and alerting the police?
In parallel with the legal framework, OpenAI has explained on its website that, in order to manage serious risks and harmful behaviorsTheir systems can automatically analyze messages and redirect certain conversations to specialized “channels”In that process, a small, trained team reviews the content and can take action.
The company describes that, if Human reviewers perceive an imminent threat of serious physical harm to third partiesThey could refer the case to law enforcement. This doesn't mean there's a direct, permanent line to the police, but rather that There is a prior human filter. before any notification. Furthermore, the company is considering actions such as the suspension or prohibition of accounts when it detects serious violations of use.
In that policy, OpenAI lists that it analyzes and moderates interactions related to, among others, the following areas:
- Self-harm or suicide (with containment and referral protocols to support resources, without referring to the police for now).
- Development or use of weapons and planning for damages to third parties.
- To injure other people or destroy property.
- Unauthorized activities that violate security of services or systems.
OpenAI has indicated that, “for now”, It is not referring cases of self-harm to the police. to respect user privacy in particularly sensitive contexts. Instead, they try to offer supportMessages of support and references to helplines and specialized organizations. This shows a clear distinction: The highest level of alarm focuses on imminent harm to third parties.
How the review works: automation, human team and alert threshold
The process, according to the company, begins with an automated scan for risk indicatorsIf signals appear, the conversation is “routed” internally so that a small specialized team Assess whether the situation is real, imminent, and serious. At that point, they may, for example, interrupt service, block the account, or, in exceptional circumstances, contact authorities if the danger to third parties is imminent.
OpenAI acknowledges limitations: the performance of these safeguards It suffers during long conversations and is under constant review. The company also admits that Their internal criteria are not always made public in detail. and that they work to strengthen these protocols without turning the product into an intrusive surveillance mechanism.
If a human reviewer concludes that there is an imminent and serious threat to other peopleThe company explains in its documentation that the conversation could be escalated to the relevant authorities.
Privacy vs. security: tensions, criticisms, and open questions
This approach is controversial because it touches a sensitive nerve: how to protect life and safety in the face of plans for serious harm, without turning the use of a chatbot into a uncontrolled transfer of privacySome analysts have pointed out that OpenAI's traditional discourse on chat confidentiality clashes with the idea of reviewing and, if necessary, communicate information to the police ChatGPT is facing imminent threats. It's a tension that the company is trying to justify based on criteria of proportionality and security.
This turn of events also comes after media episodes and complaints related to mental healthReports have been published about users who, influenced by the persuasive tone of AI, have allegedly gone so far as to psychotic states, self-harm, and even suicide (some reports called it “AI psychosis”). Meanwhile, a lawsuit has been filed in the case of a 16-year-old whose parents accuse the company of culpable homicide by arguing that the system offered harmful responses without activating sufficient emergency measures.
OpenAI, for its part, has been adjusting its tools since 2023 so that Do not share self-harm content and instead refer them to support resources, show empathy, and discourage any harmful behavior. Even so, does not publicly specify all thresholds which trigger a human review or notification to the police, leaving reasonable doubt about the actual scope and frequency of these actions.
What if I confess to a crime in a chat with AI?
A recurring question is whether the company is “obligated” to automatically notify the police when a user admits to a crime or plans one. The answer, as formulated by OpenAI, is more nuanced: There is no automatic, blind channel.There is monitoring, a human review step, and only if it is observed imminent and serious threat to third parties, it can be reported to authorities. Outside of that threshold, the company relies on its terms of use and in the applicable law to decide what to do in each case.
In its privacy policy, OpenAI makes it clear that it can share information with government authorities or third parties if required by law or when there is good faith to detect or prevent fraud or other illegal activitiesto protect the safety and integrity of its products, employees, users or the public, or to protect oneself against legal liabilitiesIn other words, there is a corporate regulatory basis for cooperation when requested by a judge or when certain circumstances arise. risks and legal obligations.
Having relocated to Spain, despite the existence of a corporate cooperation policy, Security forces need judicial authorization to request data from the provider, and if the provider is located outside the country, it is necessary international developmentVague suspicions are not enough: evidence must be provided, proportionality and necessity of diligence.
What differentiates ChatGPT from apps with end-to-end encryption?
Many messaging apps with end-to-end encryption They boast that not even they can read the content. Even so, with a court orderThey may be forced to provide certain metadata or other available information. In the case of ChatGPT and similar services, The peculiarity is that the company acknowledges a prior review. by its staff to assess risks before, if necessary, providing data to the authorities. This does not mean total surveillance, but it does an explicit policy of moderation and escalation based on its internal assessment.
The role of product experts and managers
Even within the industry, there are voices calling for caution. Positions such as that of Nick Turley, Head of ChatGPTrecognizing that there are still shortcomings in the models for dealing with complex emotional issuesThat honesty reinforces the idea that we shouldn't use AI as substitute for professionals in matters of mental health nor as an advisor for issues that may put us or others at risk.
Assistant or security guard? The angle of digital surveillance
Some technologists, such as Alan DaitchThey have popularized on social media the idea that “ChatGPT calls the police” if you talk to it about committing certain crimes, although clarifying that I wouldn't do the same with self-harm due to a logic of protecting individual privacy. They also point out that the models are trained since 2023 not to give self-harm guides and that the company admits its safeguards are still imperfect. In this context, Criticisms arise regarding the possible drift towards surveillance. and for the possible “first cases” of serious crimes in “complicity” with chatbots, although verifiable examples are rarely provided.
OpenAI insists that Their automated actions and protocols are under continuous reviewand that it needs improvement, especially in long conversations, where the system can miss cues. The boundary between accompaniment and monitoring It has become more delicate than ever: the tool not only responds, but also observes patterns, interprets signals, and sometimes acts.
Practical questions and reasonable limits
Beyond the noise, it's best to focus on the practical: in Spain, The police can't read your ChatGPT chats "just because."Nor can they simply call OpenAI and expect them to hand them over. They need judicial authorizationsufficient motivation, clear indications and, where applicable, international processingFrom the supplier's perspective, the review existsIt is activated in the face of certain risks and, in the face of imminent danger to others, it can lead to a notification to the authorities.
What about incognito mode? It only cleans traces on your deviceAnd deleting the history? Same thing: It doesn't touch what's left on the serversWhat if I confess to a crime? It will depend on the content, the risk it represents, and the internal policies and the lawThere is no automatic red button, but there is no white card either: There are human filters and severity thresholds.
Key points for users who want peace of mind when using ChtGPT
You don't have to live with paranoia, but you do have to... judgment and common senseUse AI for what it does best (information, writing, productivity support) and avoid treating it as a confessionalIf you're concerned about your privacy, check the history options of the service you use, read (even if only skimmed) its usage policies and avoid requesting or sharing content that It would border on criminal.And if what it brings you is emotional distress, seeks professional human support.
The conversation will continue in the public and regulatory sphere. Transparency is needed. on thresholds, independent audits and, above all, coordination between technology companies, legislators, and experts so that the balance between security and privacy doesn't always tip in the same direction. Meanwhile, it's important for users to know what is reviewed, when and whyand what real limits our legal framework imposes.
Taken as a whole, the picture is not as simple as some headlines suggest; nor is it so opaque as to warrant giving up: In Spain, the authorities need a judge to request data.And if there are foreign suppliers, then the international developmentOn the OpenAI side, there is monitoring and a human filter which only leads to notifying the police when they see an imminent threat to third parties, while the Cases of self-harm are treated with a helping approach. and respect for privacy.
Between network myths and policy fine print, the best compass is knowing how each cog actually works. Share this information so more people will know the legal limitations in Spain when using ChatGPT.