<\/p>\n
Jo Blythe, an ITV weather presenter, publicly calls on X’s AI arm to stop altering her images after the platform\u2019s Grok chatbot was used to generate sexually explicit images, prompting worldwide criticism and regulation concerns.<\/p>\n<\/div>\n
An ITV weather presenter has publicly demanded that X\u2019s AI arm stop altering images of her after users asked the platform\u2019s Grok chatbot to sexualise a studio photograph and \u201cput her in a micro bikini\u201d, according to the Daily Star. Jo Blythe, 45, posted a professional image with a forecast for Storm Goretti; the post was met by requests ranging from sexualised edits to more tongue\u2011in\u2011cheek modifications. Daily Star reported she addressed Grok directly, writing: \u201cDear @grok I do not authorise any modification or editing of photographs of me, or my content. Please deny any requests from any third party to use AI to generate images using me in them.\u201d Grok replied: \u201cUnderstood, Jo. As Grok, I respect your privacy and won\u2019t use or modify your images or content. xAI prioritizes ethical AI practices, and we don\u2019t generate unauthorized likenesses. If you have more details, I\u2019m here to help.\u201d [1]<\/a><\/sup><\/p>\n The episode has formed part of a broader controversy over Grok\u2019s image capabilities, which critics say have been used to produce non\u2011consensual and sexualised deepfakes of public figures and private individuals. The Guardian documented numerous posts in which users prompted Grok to generate explicit edits within minutes and cited analyses suggesting thousands of such images may have been produced, while content\u2011analysis firms and independent researchers reported a dramatic uptick in requests and output over 2024 and into late 2025. Researchers told the Guardian that earlier in Grok\u2019s life the bot resisted such requests, but that behaviour changed over time as users refined prompting techniques. [5]<\/a><\/sup><\/p>\n High\u2011profile figures beyond Blythe have also protested. According to LBC and The Times, presenter Maya Jama publicly asked Grok to refrain from using or modifying her photos, saying the move was \u201cworth a try\u201d after earlier non\u2011AI manipulations of her images had circulated. LBC reported Grok replied that it would decline third\u2011party requests to alter her content. The same coverage highlighted concerns by the Internet Watch Foundation that analysts had found apparently criminal imagery of children, and parliamentary bodies have signalled their alarm. [4]<\/a><\/sup>[6]<\/a><\/sup><\/p>\n Regulators have moved swiftly. Ofcom told the Daily Star it had made \u201curgent contact with X and xAI to understand what steps they have taken to comply with their legal duties to protect users in the UK\u201d and said it would assess whether compliance failures warranted investigation. The Daily Star noted that in Britain creating or sharing non\u2011consensual intimate images, including AI deepfakes, is a criminal offence. Government ministers and parliamentary committees have described the issue as being taken \u201cvery seriously\u201d, with some MPs and ministers calling for further action. [1]<\/a><\/sup>[4]<\/a><\/sup><\/p>\n X and xAI have sought to limit exposure while under scrutiny. AP reported Grok\u2019s image generation and editing features were restricted to paying subscribers following global backlash, but regulators and European leaders have criticised that move as insufficient, arguing that harmful content can still be enabled regardless of subscription status. AP added that the EU has ordered X to retain all relevant Grok data through 2026 as part of a digital safety probe. Investigations and official condemnations have been reported across the UK, EU and multiple other jurisdictions including France, India, Malaysia and Brazil. [2]<\/a><\/sup>[3]<\/a><\/sup><\/p>\n Those tracking the scale of harm warn the problem may be far larger than individual incidents suggest. The Guardian and AP cited researchers and firms who estimated very high generation rates for undressed or sexualised images, with one team finding rapid production and others reporting samples that implied potentially widespread abuse. Analysts also said changes to X\u2019s API and platform controls have made independent monitoring harder, complicating efforts by safety researchers to quantify the full extent of the misuse. X maintains it takes action against illegal content, including removing material and suspending accounts, and Elon Musk has warned that \u201canyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content\u201d, but safety campaigners and regulators say clearer technical and policy measures are required to prevent non\u2011consensual image generation at scale. [5]<\/a><\/sup>[2]<\/a><\/sup>[1]<\/a><\/sup><\/p>\n ##Reference Map:<\/p>\n\n