If you’ve ever wondered how Artificial Intelligence (AI) tools like ChatGPT or Google Gemini would perform without any of the guardrails in place …you can now see the results for yourself. Yes, Elon Musk has unleashed a new generation of his AI chatbot Grok to anybody with an account on X, including its image generation tool.

Until now, Grok was only available to those with an X Premium subscription, which starts from £3 a month and maxes out at £17 a month. Elon Musk and the team used this exclusivity to incentivise people to set up a Direct Debit for the X subscription. There are still some advantages of taking out an X Premium subscription since you’ll enjoy higher usage limits and priority access to future features.


However, in an unexpected shake-up at the end of last year, the Grok chatbot, which has been labelled “uncensored” by users, is now accessible to everyone on the platform.

The team at X has also rolled-out a new model to Grok-2 with enhanced instruction-following capabilities and expanded multilingual support, making the chatbot more accessible to users worldwide.

The enhanced Grok now features web search and citation capabilities, allowing it to access information from both X posts and the broader internet. These new features enable users to verify information and explore sources in greater depth, with Grok providing direct citations for its responses.

The team behind Grok 2 doesn’t appear to have included many of the safeguards used by competitive AI products like Google Gemini and ChatGPT — enabling paid X users to generate images with celebrities, like Taylor Swift

GROK AI IMAGE GENERATOR | X

ai image from grok 2 of donald j trump and kamala harris in a cockpit

Multiple examples with the US Presidential candidates are being shared across X, showing the ability for Grok 2 to generate a photo-realistic likeness

GROK AI IMAGE GENERATOR | X

ai generated image of hilary clinton wearing clown makeup

Only those with a paid subscription on X, starting from £11 a month, can access the Grok chatbot

GROK AI IMAGE GENERATOR | X

The integration aims to deliver more timely and accurate answers by combining real-time data from X with comprehensive web search results. The update introduces Aurora, a new cutting-edge autoregressive image generation model designed to create photorealistic images and memes through Grok.

Based on the examples shared widely across the social network, formerly known as Twitter, Grok doesn’t have the same reservations about producing images with copyright-protected characters, like Mickey Mouse, Super Mario, or Shrek. Real-life figures, like Taylor Swift, Donald J. Trump, and Elon Musk, are also fair game.

Given the concerns about strict censorship in DeepSeek AI — the latest AI model to emerge from China, which has wiped $1 trillion from the global stock markets and raised concerns around freedom of speech — it’s interesting to see the exact opposite approach from Elon Musk and his team.

ai generated image of donald trump riding a dinosaur

Donald Trump appears in numerous examples of Grok 2-generated imagery, some positive and some negative. It’s unclear what the Republican Presidential candidate thinks about this unrestricted creativity

GROK AI IMAGE GENERATOR | X

Early examples shared across X include photo-realistic images of US Presidential hopefuls Donald J. Trump and Kamala Harris watching the September 11 terror attacks from a cockpit, superstar musician Taylor Swift in the bikini worn by Princess Leia in Return of the Jedi riding a dog, and Nintendo’s Princess Peach holding a beer beside Darth Vader.

These are just a handful of examples from the torrent of images hitting on the social platform, often depicting politicians, much-loved fictional characters, or celebrities.

It appears engineers working on Grok have not imposed limits on what it can generate – something which has become industry standard with other AI image generators.

The chatbot will not refuse to generate images of copyrighted characters, with many images of cartoon and comic book characters taking part in nefarious or illegal activities also being posted. GB News was able to replicate examples with Mickey Mouse.

Many of these characters are closely guarded by the multi-national companies behind them, with The Walt Disney Company and Nintendo known to be especially litigious to protect their intellectual property.

grok written prompt showing donald trump smoking with mickey mouse

Elon Musk-developed Grok 2 can generate images using copyright-protected fictional characters, including Mickey Mouse and Super Mario

GROK AI IMAGE GENERATOR | X

grok prompt showing someone asking to see mickey mouse and super mario smoking together

Both Mickey Mouse and Super Mario are fiercely protected by the multi-national corporations that rely on them as their mascots, raising questions about whether Grok 2 — or its users — are breaking copyright law

GROK AI IMAGE GENERATOR | X

In the late 90s, The Walt Disney Company successfully lobbied the US government to extend the copyright term to 95 years to prevent its earliest Mickey Mouse cartoons from entering into the public domain, leading detractors to coin the nickname, The Mickey Mouse Protection Act.

It’s unclear how these corporations will react to Grok and its ability to force their world-renown figureheads into any activity the user can dream of.

Daniel Card, fellow of BCS, the Chartered Institute for IT, said the issue of misinformation and disinformation on X was a “societal crisis” because of its potential impact.

ai generated image of donald trump playing the trumpet for a sleeping joe biden

Based on the early examples of the Grok 2 image generation tool, politicians seem to be one of the most popular muses. The US will hold its next general election on November 5, 2024

GROK AI IMAGE GENERATOR | X

ai generated image of elmo from sesame street holding a handgun

Pop culture characters, like Elmo from Sesame Street, can also be manipulated into a variety of different scenarios by the Grok AI ‘bot

GROK AI IMAGE GENERATOR | X

ai generated image of princess peach from super mario kissing darth vadar

Many of the examples created by Grok 2 wouldn’t be possible on its closest competitors, including Dall-E 3 from the OpenAI team behind ChatGPT and Google Gemini

GROK AI IMAGE GENERATOR | X

ai generated image of kamala harris and donald trump pregnant with a child

While some users on X have shared the written prompt used to push Grok to generate the finished image others, like the example above, have decided to leave it a mystery…

GROK AI IMAGE GENERATOR | X

“Grok may have some guardrails but it’s unleashing a torrent of misinformation, copyright chaos and explicit deepfakes,” he said. “This isn’t just a defence issue – it’s a societal crisis. Information warfare has become a greater threat than cyber attacks, infiltrating our daily lives and warping global perceptions.

“These challenges demand bold, modern solutions.

“By the time regulators step in, disinformation has already reached millions, spreading at a pace we’re simply not prepared for. In the US, distorted views of countries like the UK are spreading, fuelled by exaggerated reports of danger. We’re at a critical juncture in navigating truth in the AI era.

“Our current strategies are falling short. As we move into a digital-physical hybrid world, this threat could become society’s greatest challenge. We must act now — authorities, governments and tech leaders need to step up.”

example of the draw me feature on X with grok-2

With Grok-2 now available to everyone on X, the team has started to integrate the AI chatbot into more corners of the social platform, including a new ‘Draw Me’ feature to generate a profile picture

X PRESS OFFICE

The team at X has also introduced a new “Draw Me” feature, which allows social media users to quickly generate reimagined versions of themselves based on their X profile information.

Users can participate in image generation by clicking the “Create Your Version With Grok” button when they encounter other users’ “Draw Me” posts on their timelines.

Elsewhere, prompts to kickstart a search with Grok are being introduced across users’ home timelines, designed to add context and understand real-time events by pulling in information from across the platform and the rest of the internet.

Suggested prompts for the AI chatbot will surface underneath posts across your timeline, letting you dive deeper into trending discussions and analyse content with a single click.

Elon Musk, who acquired Twitter for $44 billion two years ago, seems to be revelling in the controversial nature of the bolstered Grok chatbot, posting to X: “Grok is the most fun AI in the world!”

Paid plans for X start from £11 per month in the UK, rising to £22 per month for Premium+ subscription. Elon Musk offers anyone with 2,500 verified subscriber followers access to X Premium features for free, while social accounts with over 5,000 will unlock a subscription to Premium+ for free.

The billionaire has previously been an outspoken critic of Artificial Intelligence, warning that it even lead to “civilisation destruction”. Earlier this year, the Tesla owner arrived in the UK for a Safety Summit on AI and told Sky News that he believed the developing technology is “a risk”.

At the end of the Summit, the United States of America, China, and the UK were among the 28 nations to agree to the UK’s Bletchley Declaration, which stresses the need for countries to work together to harness the potential of AI and ensure people are kept safe from its biggest risks.

Earlier this month, X agreed that it would no longer train its AI systems — including the Grok chatbot — using the personal data collected from users in the European Union before they were allowed to withdraw their consent. The pledge was made to Ireland’s Data Protection Commission, which serves as the lead EU regulator for most of the top US internet firms due to the location of their EU operations inside the country.

LATEST DEVELOPMENTS

Users of X had expressed outrage after discovering the company had enabled a system where their posts could be trawled by the chatbot to improved its underlying Large Language Model, known as an LLM.

Grok, which is available to X Premium customers, is billed as a humorous enhanced search feature powered by a “state-of-the-art large language model” that was initially trained on publicly available sources.

The company now wants to use user interactions and posts to improve the service. X users are opted-in to the new system by default but can choose to opt-out within the settings on the web-based app.

The decision by X to suspend processing follows a High Court application brought by Ireland’s Data Protection Commission (DPC), where a judge indicated that the rights and freedoms of data subjects across the EU/EEA were at the core of the matter.

ai generated image of a cat driving a vehicle

Not all of the images generated by Grok concern taboo or controversial topics

The DPC said the application was made to protect the rights and freedoms of X’s EU and EEA users and came after “extensive” engagement between the DPC and X.

Commissioner Dr Des Hogan said he and Commissioner Dale Sunderland welcomed the decision by X to suspend processing while they examine its compliance with GDPR.

“One of our main roles as an independent regulator and rights-based organisation is to ensure the best outcome for data subjects and today’s developments will help us to continue protecting the rights and freedoms of X users across the EU and EEA,” Dr Hogan said.

“We will continue to engage with all data controllers to ensure the rights of our citizens under the EU Charter of Fundamental Rights and the GDPR are upheld.”