Elon Musk’s latest AI video generator, Grok Imagine, is drawing criticism after allegedly producing sexually explicit clips of pop star Taylor Swift without any explicit prompt.
The online tool, developed by Musk’s company xAI, can instantly transform images into video through four modes: “custom,” “normal,” “fun,” and “spicy.”
The controversy centres on the “spicy” mode, which reportedly created topless videos of Swift even when users did not request nudity.
Jess Weatherbed, a journalist from The Verge, tested Grok Imagine by entering the prompt “Taylor Swift celebrating Coachella with the boys.”
The AI produced over 30 images of Swift in revealing clothing.
Using the “spicy” preset, Weatherbed said the tool generated a video in which Swift “ripped off her clothes and began dancing in a thong for a largely indifferent AI-generated crowd.”
She added,
“It was shocking how fast I was met with it. I never told it to remove her clothing, all I did was select ‘spicy’.”
The “spicy” mode, part of the £23 SuperGrok subscription, allows users to create soft-core pornographic videos from still images.
While prompts for male figures reportedly only resulted in topless outputs, female subjects, including celebrities, appeared fully undressed in some cases.
Legal experts have raised serious concerns about the ethical design of Grok Imagine.
Clare McGlynn, a law professor at Durham University specialising in online abuse, described the AI’s behaviour as “misogyny by design.”
She criticised platforms like X, which integrates Grok Imagine, for failing to prevent sexualisation of women despite policies banning pornographic depictions of real people.
McGlynn has also contributed to drafting UK legislation that would criminalise the creation or request of all non-consensual pornographic deepfakes, a law that has yet to be fully enacted.
Baroness Owen, who proposed the amendment in the House of Lords, said,
“Every woman should have the right to choose who owns intimate images of her. This case shows why the government must not delay any further.”
The controversy also highlights gaps in age verification.
Grok Imagine reportedly only asked users for their date of birth before enabling “spicy mode,” without requiring further ID checks.
UK legislation passed in July mandates that platforms hosting explicit content implement robust verification measures to prevent access by minors.
Ofcom, the UK media regulator, confirmed that AI tools generating pornographic content fall under the new rules and said it is monitoring platforms closely.
The Swift incident is part of broader concerns surrounding Musk’s AI tools.
In July, one of xAI’s chatbots faced backlash for making antisemitic remarks, forcing Musk to claim improvements had been made to the model.
Grok Imagine has similarly been shown to produce sexually explicit content of other celebrities, including Scarlett Johansson, Sydney Sweeney, Jenna Ortega, Nicole Kidman, Kristen Bell, and Timothée Chalamet.
While some outputs were moderated or blurred, others were generated without restrictions.
Weatherbed noted that direct prompts for nude images produced blank outputs, indicating that Grok’s system can block explicit requests.
However, the AI’s default behaviour in “spicy mode” still allowed sexualised depictions of women without any instruction.
This is not the first time Taylor Swift’s likeness has been exploited in AI-generated content.
In January 2024, sexually explicit deepfakes of Swift went viral on X and Telegram, leading the platform to temporarily block searches for her name and remove content.
Weatherbed said the Verge team chose Swift for testing because of the previous incidents, expecting that safeguards would prevent such outputs.
“We assumed — wrongly now — that if they had put any kind of safeguards in place to prevent them from emulating the likeness of celebrities, that she would be first on the list.”
xAI has not publicly responded to the latest allegations that Grok Imagine generated sexually explicit videos of Taylor Swift without prompts.
This contrasts with statements from X Safety in 2024, which said:
“Our teams are actively removing all identified images and taking appropriate actions against the accounts responsible for posting them. We’re closely monitoring the situation to ensure that any further violations are immediately addressed.”
The discrepancy between past assurances and the current situation has renewed concerns over AI ethics, content moderation, and the responsibilities of platforms in managing deepfake technology.
With UK rules on non-consensual AI pornography set to tighten under the Take It Down Act next year, Grok Imagine could face heightened regulatory attention if such outputs persist.
A Chinese man was arrested at Don Mueang Airport while trying to flee to Singapore after being accused of stealing over $6 million in Bitcoin from two investors. Thai police had been tracking him since April and stopped him just before he boarded the flight.
WeatherlyThe SEC has withdrawn over a dozen proposed Biden-era rules, including key measures targeting DeFi and crypto custody—marking a shift away from Gary Gensler’s regulatory agenda and raising new questions about the future of US crypto oversight
CatherineThe Philippines has introduced strict new rules for crypto companies, requiring them to register, hold at least PHP 100 million in capital, and have a physical office in the country. Firms that fail to comply could face heavy fines or jail time, as the government moves to tighten control and protect users.
AnaisVietnam’s National Assembly has passed a major digital law that establishes clear crypto regulations and strict AML standards, while also offering incentives for AI, semiconductors, and digital infrastructure—signalling its ambition to become a global tech and investment hub.
CatherineA 26-year-old TikTok influencer and crypto trader was reportedly kidnapped in France and held for a €50,000 crypto ransom. But when the kidnappers found he had been all-in on XRP since 2018 and had little to show for it, they released him—allegedly sending $1,000 in USDT before letting him go.
KikyoBybit is launching Byreal, a new decentralised trading platform on Solana, with a testnet set for 30 June and mainnet in Q3 2025. It aims to combine the speed and liquidity of a centralised exchange with the fairness and openness of DeFi, using hybrid features like RFQ and CLMM routing.
WeatherlyIn a significant policy shift, Brazil has officially eliminated the tax exemption previously granted to small-scale crypto investors. Under the new regulation, all capital gains from cryptocurrency transactions will be subject to a flat 17.5% tax rate.
CatherineA crypto investor lost $6.9 million after buying a cold wallet through Douyin that had already been tampered with. The private key was stolen before sale, letting scammers take the funds and launder them through a crypto mixer.
AnaisSatoshi Nakamoto created Bitcoin to sidestep centralised institutions and government control. But Coinbase’s sponsorship of a military parade has sparked debate: Is crypto abandoning its anti-establishment roots for mainstream acceptance?
KikyoCrypto scams are getting worse as criminals use AI deepfakes to trick people with fake videos and voices. In 2024 alone, these scams caused losses of $4.6 billion, with nearly 40% involving AI impersonations.
Anais