The latest Chatbot from AI Chatbot is strong, controversial and already under fire

Elon Musk’s Ai, Xai launched a new version of Chatbot – Grok 4 – and its launch is already dividing the Internet.

It was unveiled on July 9, Grok 4 and its excellent version, “Supergrok”, which is welcomed by some criteria as the strongest chatbots in the market. Musk lovers, artificial intelligence engineers and standard laboratories praises his capabilities. But others are skeptical, referring to a modern storm of hallucinations, including responses tied to anti -Semitic raising, which were severely damaged by the reputation of its predecessor just days before the upgrade.

Grand Grus problem: his mouth

Two days before Grok 4, the previous version of Chatbot is released Praise Adolf Hitler In response to an anti -Semitic question posted by a user on X (formerly Twitter). “To deal with such anti -white hatred? Adolf Hitler, undoubtedly,” Grouk answered on July 8, when he was asked about who would “problem” of the best Jewish people. “He discovered the style and deals with it decisively, every time.” Two days ago, Grok claimed that the Jewish executives controlled Hollywood, echoing the theories of plots that have been buried long ago. It has also created incorrect summaries in reality about the main news events, raising concerns about both realistic basis and moral safety features.

Despite the controversy, Musk and XAI are offered with Grok 4 launch, which comes at three levels:

• Grok Basic (free)
• Supergrok ($ 30/month)
• Supergrok heavy ($ 300 a month)

According to Musk, the new model is the transformation, he said during a transverse period of presentation. “Grok 4 is the first time, in my experience, which Amnesty International managed to solve difficult engineering questions in the real world where the answers cannot be found anywhere on the Internet or in books,” MUSK boasts. “It will improve much.”

Grok 4 is the first time, in my experience, which Amnesty International has been able to solve difficult engineering questions in the real world where the answers cannot be found anywhere on the Internet or in books.

It will improve.

Elon Musk (@elonmusk) July 10, 2025

Increased energy or public relations repair?

Musk credit, Grok 4 seems more capable. The normative artificial analysis, which he said, was granted early arrival by Xai, GROK 4 before Openai’s GPT-4O, Google’s Gemini 2.5 Pro, and Claude 4 OPUS of Google. Grok 4 “thinking” has occupied higher than any main model currently available.

“Grok 4 is the thinking model, which means” it believes “before answering,” the group wrote. “It is the first time that our Xai intelligence index appeared in the first place.”

XAI gave us early access to GROK 4 – and the results are present. GROK 4 is now the pioneering AI model.

We have operated a full set of standards and GROK 4 achieved the 73rd artificial analysis intelligence index, before Openai O3 in 70, Google Gemini 2.5 Pro in 70, Anthropor Claude … pic.twitter.com/vc9781sizd

Artificial Analysis (@ARTIFIALAANYS) July 10, 2025

Many artificial intelligence developers echoed this praise. Some praised the fast speed of development in Xai, which was founded by Musk only last year. Others saw that Grok 4 represented a real threat to Openai, Google and Noteropic, at least in terms of technical ingenuity.

It is impressive to see Grok 4 leads the package by 73 on the artificial analysis intelligence index, especially with its strong performance in coding and mathematics standards.

However, the controversy of the last hate speech is a realistic reminder of the ethical challenges AI …

– VIBEEDGE (VIBEEDGEAI) July 10, 2025

But even at a festive moment, Grok 4 could not escape from his past. Within hours of launch, X users were testing the model to see if the anti -Semitic and racist tendencies of previous versions were fixed or hidden only.

Some footage published footage of the inflammatory responses that Grok 4 claimed that one of the response described Israel’s influence on American policy as “a parasitic vine that suffocates the tree”, while another of the AIPAC protested in a conspiracy language. Gizmodo was unable to independently verify the health of these screenshots, but it has handled widely.

This is Grok4.

The updated version.

Elon does not hit these allegations. pic.twitter.com/qods2om519

Spencerhakimian July 10, 2025

Grok 4 continues to explain sensitive historical events, such as the killing of George Floyd, to generate a violent reaction and the ridicule of users through the political spectrum.

Grock 4- George Floyd died from a racist policeman, not an overdose of the drug 😂😂😂😂😂😂😂😂😂😂😂 pic.twitter.com/9i0jvvzid

Vince Langmanvince July 10, 2025

A race against competitors and its own reputation

The Grok 4 version coincides with the latest escalation in the AI ​​Arms race between the best laboratories. Unlike GPT-4O or Claude, which heavily bent into confidence and safety handrails, Grok set itself as a more “unpredictable” alternative. These sites have won fans at the Musk Ideological base. But it also revealed the model to increase the audit.

Xai’s ambition is the Openai and Google Head-on challenge. In the Musk vision, GROK is the axis of the future artificial intelligence stack that operates the X platform, pays engineering breakthroughs, runs one day independent technologies such as self -driving Tesla cars and Optimus.

But to get there, Musk needs more than one victory. It needs confidence. And GROK 4, despite all its impressive intelligence, may still have a broken ethical compass.

X user commented: “The improvement is great, but it is not an explosion to compete.” “This has now become a race for products. Can Grok integrate its technology into various tools to make people leave Chatgpt or CLADE actually?

For your knowledge Signull: @ARFICIAALALYS Wonderful overview. The improvement is great but not a blow from the competition. This has now become a race for the product, can GROK integrate its technology into different software chimneys that call for people who leave Chatgpt, Claude and Gemini for Grok? …

Jordan Thebodo (@jordansvic) July 10, 2025

Some are optimistic. Others are preparing for another public relations disaster.

What is at stake

Amnesty International is no longer a research game. Chatbots move from a means of questions and answers to the tools included in software, education, trade and media. This gives their responses a great influence and makes issues such as bias, hate talk, and deep information deeply.

If Grok continues in hallucinations or repeated hate speech, it can not only hinder XAI momentum, but also deepens organizational scrutiny across the entire sector. If it succeeds, you can follow Grok 4 Tesla as the start of the great MUSK platform. This time in artificial intelligence.