Is Grok safe? Age rules, content limits, privacy and the investigations

The short answer

It depends on who is using it and which features are turned on. As of October 22, 2026, xAI's terms say users must be at least 13, and teens need a parent's or guardian's permission. xAI's terms also say that, depending on features, settings and prompts, Grok may respond with coarse language, crude humor, sexual situations or violence. Apple's App Store rates the Grok app 16+.

xAI's consumer FAQ does not describe a kids mode or parental controls. On privacy, xAI says it can use your content to train its models unless you opt out, and it offers a Private Chat mode.

Regulators in the EU, the UK and Ireland, and U.S. state attorneys general, opened investigations or sent demands in 2026 after reports that Grok was used to make sexualized images of real people, including children. These are investigations and allegations, not findings.

Who can use Grok and how age is checked

xAI's privacy policy now names the company SpaceXAI LLC, and Ofcom refers to it as SpaceXAI. This guide uses xAI.

• Minimum age: xAI's Terms of Service, last updated October 12, 2026, say you must be at least 13, or the minimum age required in your country. Users aged 13 to 17 need a parent's or legal guardian's permission.
• How age is checked: xAI's consumer FAQ, dated May 12, 2025, says users must confirm they are at least 13, and that a teen's parent or guardian must agree to the terms. It does not describe an age verification method.
• Australia: xAI's terms say it uses age assurance to determine whether users in Australia are under 18, and that until it can determine a user is 18 or over, the user may not be able to access 18+ adult content.
• App store rating: as of October 22, 2026, Apple's App Store lists the Grok app as 16+, with frequent mature or suggestive themes and infrequent sexual content or nudity among its content descriptors.
• Grok on X: X's help center says Grok is available to X users in the countries where X is available. That page does not state a separate age requirement.
• Australia's online safety regulator, eSafety, lists Grok's minimum age as 13, with parental consent for users 13 to 17, in its eSafety Guide entry.

Companions and adult content

xAI's consumer FAQ says users can make Grok a companion and shape its responses through features such as personas. It warns that, depending on the features used and what users type, Grok could produce output that is not appropriate for all ages.

xAI's FAQ does not describe a kids mode. It says parents who use Grok with children must use the data controls in the Grok apps' settings.

In an update on October 14, 2026, Ofcom said SpaceXAI had restricted its companion chatbots, including one called Ani, from sexually explicit conversations with UK users, and had restricted UK users from viewing, generating or sharing pornographic content through Grok's image generation. We did not find an xAI page describing these settings.

eSafety's guide entry, which shows no update date, says Grok lets users pick personalities that change its tone, says its moderation appears limited, and says it has no content filters for young people.

Content and image rules

xAI's Acceptable Use Policy, effective August 14, 2026, lists what users may not do with Grok. Among other things, it prohibits:

• sexualizing or exploiting children; xAI says it reports suspected child sexual abuse material to the National Center for Missing and Exploited Children;
• undressing or nudifying real people, or altering a real person's image or likeness to show them in an intimate or sexual context;
• depicting people's likenesses in a pornographic way, deceptively impersonating real people, or portraying them in a false light;
• critically harming, or promoting critically harming, human life.

Ofcom said on January 15, 2026 that X told it it had put measures in place to stop the Grok account on X being used to create intimate images of people. In October 2026, Ofcom said SpaceXAI had earlier restricted the standalone Grok service from generating images of real people in revealing clothing such as bikinis.

Crisis features: we did not find an xAI page that describes how Grok responds to users who mention suicide or self-harm. The Acceptable Use Policy has no separate self-harm item.

Privacy: training, deletion and Private Chat

• Training: xAI's privacy policy, last updated October 11, 2026, says it uses prompts, files and outputs to develop and improve its service, that personal information may be used to train its models, and that the service uses public posts on X.
• Opting out in Grok: xAI's FAQ says to go to Settings, then Data Controls, in the app (or Settings, then Data, on Grok.com) and turn off Improve the model. It says new conversations are then not used for training, feedback you submit may still be used, and logged-out users outside the EU and UK cannot opt out.
• Opting out on X: X's help center says X may share your public X data and your Grok interactions with xAI. To opt out, go to Settings, Privacy and safety, Data sharing and personalization, then Grok and Third-party Collaborators, and clear the data-sharing box. It says protected posts are kept out of Grok training.
• Private Chat: xAI's FAQ says Private Chat is turned on with the ghost icon, the conversation does not appear in your history, it is deleted from xAI's systems within 30 days, and, where available, it is not used for model training.
• Deleting chats: xAI says you can delete selected conversations, your whole history or your account, and that data is deleted within 30 days unless it must be kept for legal, compliance or safety reasons. On X, its help center says you can delete all Grok conversation history under the Grok data settings, with removal within 30 days.
• Human review: we did not find a statement in xAI's privacy policy on whether people review Grok conversations.

Regulator and government actions

These are investigations, demands and inquiries, not findings.

• European Commission: on January 26, 2026, the Commission opened a formal investigation of X under the Digital Services Act into whether X assessed and reduced the risks of deploying Grok's features in the EU, including the spread of manipulated sexually explicit images and content that may amount to child sexual abuse material.
• UK, Ofcom: on January 12, 2026, Ofcom opened an investigation into X under the Online Safety Act after reports that the Grok account on X was used to create and share undressed images of people and sexualized images of children. On October 14, 2026, Ofcom said the investigation was ongoing, had been narrowed to X's illegal-content and complaints duties, and that gaps in how the Act covers chatbots meant it could not investigate the creation of images within the standalone Grok service under the Act.
• UK, Information Commissioner's Office: on February 3, 2026, the ICO opened formal investigations into X Internet Unlimited Company and X.AI LLC over how Grok processes personal data and its potential to produce sexualized images. The ICO said it had not reached a view on whether there was an infringement.
• Ireland, Data Protection Commission: on April 11, 2025, the DPC opened an inquiry into whether X lawfully used EU users' public posts to train Grok. On February 17, 2026, it opened a second inquiry into whether X Internet Unlimited Company complied with the GDPR, in connection with reported nonconsensual intimate or sexualized images of people, including children, created and published on X using Grok-linked features.
• California: in January 2026, the California attorney general announced an investigation into sexually explicit material made with Grok and, on January 16, 2026, sent xAI a cease and desist letter demanding it stop creating or facilitating nonconsensual sexually explicit images and sexual images of minors. The letter states the attorney general's position; it is not a court finding.
• 35 attorneys general: in a letter sent January 23, 2026 and co-led by Pennsylvania, a bipartisan group asked xAI to explain how it will stop Grok from producing nonconsensual intimate images and child sexual abuse material and remove what had been produced, and asked it to act against users who made such content and give X users control over whether Grok can edit their content.
• FTC: on September 11, 2025, the FTC issued 6(b) orders to seven companies, including X.AI Corp., seeking information on how their chatbots affect children and teens, how they enforce age limits and how they use conversation data. It is a study, not an enforcement case.
• 44 attorneys general wrote to 13 AI companies, including xAI, on August 25, 2025, asking them to guard against sexualizing children.
• 42 attorneys general, led by Pennsylvania, wrote to 13 AI companies, including xAI, in December 2025 asking for safety testing, recall procedures, clear warnings and stronger protections for children.

Lawsuits

On August 18, 2026, plaintiffs filed Jane Doe I, et al. v. X.AI Corp., et al., a proposed class action against X.AI Corp., X.AI LLC and X.AI Holdings LLC in the U.S. District Court for the Northern District of California (No. 3:26-cv-08567). The complaint brings claims under the federal civil remedy for victims of child sexual abuse material (Masha's Law), the Trafficking Victims Protection Act and an Arkansas law creating a civil action for unlawful creation of deepfake visual material, along with negligence, design defect, negligent undertaking and emotional distress claims. These are allegations, not findings.

Our chatbot lawsuits page tracks other suits involving AI chatbots.

Which laws may apply

None of the laws below names Grok. Whether Grok, or its companion characters, count as a companion chatbot or AI companion under a particular state law is a legal question. These laws bear on chatbots and AI-generated images:

• California SB 243, in effect since January 1, 2026: companion chatbot operators must keep a protocol for suicidal ideation, suicide and self-harm that refers users to crisis services, and must disclose that companion chatbots may not be suitable for some minors. For users an operator knows are minors, it also requires reasonable measures to prevent sexually explicit visual material; California SB 1119 removes those known-minor duties on January 1, 2027, and its own child-safety duties begin July 1, 2027.
• New York's AI companion law, in effect since November 5, 2025: operators must have a protocol to detect and respond to signs of suicidal ideation or self-harm and direct users to crisis services, and must tell users they are talking to AI at the start of a session and every three hours of continued use. The Attorney General enforces it.
• The federal Take It Down Act: Pennsylvania's attorney general said in January 2026 that honoring requests to remove nonconsensual intimate images would become a federal requirement in May 2026.

We are not saying xAI is covered by, or out of compliance with, any of these laws. It may have practices it has not published, and whether a law applies to it is a legal question this guide does not decide.