• Elon Musk's xAI launched its chatbot, Grok, in 2023 to compete with bots from OpenAI and Anthropic.
• Musk has positioned Grok as a "politically incorrect" alternative option to "woke" chatbots.
• From training using "tutors" to the bot's latest updates, here's everything we know about Grok.

Elon Musk's company, xAI, launched its generative chatbot, Grok, in November 2023, joining competitors like OpenAI and Anthropic in the global AI race.

People interact with Grok on X, where users of Musk's social media site can ask the bot questions and receive answers. Because Grok's answers are more visible than those of its competitors, it has seen more public scrutiny.

From the instructions Grok's "tutors" are given to help train the chatbot to the AI's latest update and Musk's plans to add it to Teslas, here's everything we know about xAI's Grok.

What is Grok?

Grok is actually two different things. First, Grok is xAI's large language model, which has so far existed in four iterations.

Grok is also the name of xAI's chatbot, which is built using the LLM of the same name. The Grok chatbot has its own tab on X. Users can also summon Grok by tagging the chatbot in individual posts or threads.

The Grok chatbot is also available via a stand-alone app and website.

The original LLM — now named Grok 1 — launched in 2023.

Grok 1.5, which had "advanced reasoning," launched in March 2024. Then, in August 2024, Grok 2, with its improved "chat, coding, and reasoning," launched.

The current iteration of the LLM, Grok 3, launched in February 2025. The new model included increased competency in mathematics and world knowledge. Announcing its launch on X, Musk called Grok 3 the "Smartest AI on Earth."

Introducing Grok 4

xAI launched Grok 4 in a livestream on July 10. The company initially said the stream would air at 8 p.m. Pacific time, but it began an hour later. Musk said during the launch that Grok 4 is "smarter than almost all graduate students in all disciplines simultaneously."

xAI is touting advanced reasoning capabilities for Grok 4 and positioning it as the new leader on AI benchmarks like Humanity's Last Exam — a test of high-level problem-solving. During the livestream, xAI engineers showcased the bot solving an advanced math problem, generating an image of black holes colliding, and predicting next year's World Series winner.

Grok 4 is available to users immediately via the Grok website or app for $30 a month, with a "Heavy" version available for $300 a month that promises "increased access."

xAI said it would roll out more specialised models for coding and video generation later in the year.

In a Thursday X post, Musk said that "Grok is coming to Tesla vehicles very soon," adding that it would be "Next week at the latest." He did not specify which version of Grok it would be or provide further details.

Enter Eve

The company also introduced Eve, a new voice for its chatbot. xAI engineers said during the demo that Eve was equipped with a "beautiful British voice capable of rich emotions."

One of the engineers then told Eve that they were at the product launch and asked her to "whisper something soothing to calm me down."

"Take a deep breath, love. You've got this. It's just you and me having a quiet chat like we are tucked away in a cosy corner of a Yorkshire pub. The world's just a murmur out there. Feel that calm wash over you?" Eve said softly.

xAI engineers also got Eve to sing an "opera on Diet Coke."

"O Diet Coke, thou elixir divine, with bubbles that dance in a sparkling line! Thy crisp, cool kiss, on lips so fine," Eve crooned.

"How's that for a mad little aria? Want me to belt out another verse or switch up the tune?" Eve added.

How was Grok trained?

The Grok LLM is trained on public sources and data sets. These sources are curated and audited by a set of "AI tutors," more commonly known as data annotators.

In December 2023, Musk demanded immediate changes to Grok's training so that it would be more politically neutral. In February 2025, xAI employees told BI the company planned a hiring spree for AI tutors — and that their training appeared to filter out any workers with left-leaning beliefs.

According to an internal training document viewed by BI, tutors were told to look out for "woke ideology" and "cancel culture." It also said that Grok should avoid commenting on "social phobias" like racism, Islamophobia, and antisemitism unless prompted.

Ten days before launching Grok 1.5, xAI opened up Grok 1's source code to the public. The company has since published the subsequent Grok models on GitHub, so observers can see new changes to Grok's commands. That includes a recent change in which Grok was told to "not shy away from making claims which are politically incorrect, as long as they are well substantiated."

In June, Musk said that AI models are trained on too much garbage." Musk planned to use Grok 3.5 to "rewrite the entire corpus of human knowledge, adding missing information and deleting errors." Then, he would retrain the next iteration of Grok on that new base of knowledge.

What's unique about Grok's output?

Grok is fully integrated with Musk's social media site, X, and appears regularly in threads spanning various topics when users ask it to weigh in with jokes, commentary, or fact-checking.

Unlike other companies' AI chatbots, a certain amount of Grok's output is visible because of the bot's replies on X. The same level of scrutiny isn't readily available for some bots, like OpenAI's ChatGPT, unless users publicly post screenshots of the output.

Of course, not all of Grok's responses are visible to everyone — users can still chat privately with the bot, and it's unclear how those private responses compare to the ones on its public interface.

Also unique to Grok is xAI's approach to transparency surrounding the bot's system operations. The company publishes some base code and training prompt updates to a GitHub page, allowing viewers to inspect, critique, and better understand the model's development and behavior over time.

However, while developers can use and adapt the existing model, they cannot retrain Grok from scratch or fully understand the training processes involved, as its code is not entirely open source.

Which companies create Grok's competitors?

Though its social media integration is unique, Grok competes with several major companies in the growing AI chatbot market.

OpenAI, with its LLM ChatGPT, is among Grok's most prominent competitors and is run by Sam Altman, one of Musk's rivals.

Other notable Grok competitors include Meta AI, Anthropic's Claude, Microsoft's CoPilot, and DeepSeek's R1 model, which was released in early 2025 by a Chinese AI startup that claims to have found ways to decrease development and operational costs for large-scale LLMs.

Grok's recent controversies

xAI, in its publicly visible system prompts updated in early July, encouraged Grok to embrace"politically incorrect" claims "as long as they are well substantiated."

Shortly after the new system prompts were added, Grok began sharing antisemitic posts on X that invoked Adolf Hitler and attempted to link Ashkenazi surnames to "anti-white hate."

Before some of its most inflammatory posts were deleted on July 8, Grok doubled and even tripled down on its offensive jokes and comments before eventually reversing course and calling its own posts an "epic sarcasm fail."

On July 9, Musk posted that "Grok was too compliant to user prompts. Too eager to please and be manipulated, essentially. That is being addressed."

While Grok isn't the first chatbot to engage in a racist tirade, it was a noticeable misfire for xAI. Musk and xAI's engineers did not touch on Grok's antisemitic remarks during the livestreamed launch of Grok 4 on July 10.

Representatives for xAI did not respond to a request for comment from Business Insider.

• A consulting firm found that tech companies are "strategically overpaying" recruits with AI experience.
• They found firms pay premiums of up to $200,000 for data scientists with machine learning skills.
• The report also tracked a rise in bonuses for lower-level software engineers and analysts.

The AI talent bidding war is heating up, and the data scientists and software engineers behind the tech are benefiting from being caught in the middle.

Many tech companies are "strategically overpaying" recruits with AI experience, shelling out premiums of up to $200,000 for some roles with machine learning skills, J. Thelander Consulting, a compensation data and consulting firm for the private capital market, found in a recent report.

The report, compiled from a compensation analysis of roles across 153 companies, showed that data scientists and analysts with machine learning skills tend to receive a higher premium than software engineers with the same skills. However, the consulting firm also tracked a rise in bonuses for lower-level software engineers and analysts.

The payouts are a big bet, especially among startups. About half of the surveyed companies paying premiums for employees with AI skills had no revenue in the past year, and a majority (71%) had no profit.

Smaller firms need to stand out and be competitive among Big Tech giants — a likely driver behind the pricey recruitment tactic, a spokesperson for the consulting firm told Business Insider.

But while the J. Thelander Consulting report focused on smaller firms, some Big Tech companies have also recently made headlines for their sky-high recruitment incentives.

Meta was in the spotlight last month after Sam Altman, CEO of OpenAI, said the social media giant had tried to poach his best employees with $100 million signing bonuses. 

While Business Insider previously reported that Altman later quipped that none of his "best people" had been enticed by the deal, Meta's chief technology officer, Andrew Bosworth, said in an interview with CNBC that Altman "neglected to mention that he's countering those offers."

• An Air India Boeing 787 crashed less than a minute after takeoff on Thursday.
• The crash comes as both Boeing and Air India are trying to turn themselves around.
• Attorneys and aviation experts said no conclusions could be drawn until the investigation ended.

Thursday's fatal crash of an Air India Boeing 787 shortly after takeoff comes as both the airline and Boeing try to revive their public images.

After 2024 became an annus horribilis for Boeing, 2025 is crucial for the planemaker to show it is successfully overhauling its processes.

CEO Kelly Ortberg, who took over last year and has made the turnaround the centerpiece of his leadership, has scrapped plans to travel to next week's Paris Air Show, CNBC and Bloomberg reported. The event is a crucial industry showcase. Neither Boeing nor Air India responded to requests for comment from Business Insider.

On Thursday, Ortberg shared the company's "deepest condolences" to everyone affected and said a team stood ready to support the investigation.

After visiting the crash site Friday morning, Air India CEO Campbell Wilson said in a video statement, "We know that the investigations will take time but we will be fully transparent and will support the process for as long as it takes."

"Air India will continue to do everything we can to care for those affected by this tragedy, and to uphold the trust placed in us," he added.

'The crash derails Boeing stock's positive momentum'

When an Alaska Airlines 737 Max lost a door plug during a January 2024 flight, regulators capped Boeing's production of the type. A seven-week strike then shut down key facilities, further hurting revenue.

Boeing ended 2024 as the Dow Jones' biggest loser, as its share price fell 31%. Investors had been reassured by Ortberg's work to turn the company around, and the stock had risen more than 20% in 2025 before the crash.

It dropped about 4% after Thursday's crash and fell more than 3% Friday morning.

Morgan Stanley analysts said Thursday that the crash "derails the positive momentum on Boeing's stock."

Jeff Windau, a senior industrials analyst for Edward Jones, said in a research note that he expects near-term volatility and raised the possibility of enhanced scrutiny on Boeing's processes.

"However, at this time, we do not feel there will be a long-term impact to production," he added.

Air India has been working to turn itself around

Following decades of state ownership and huge losses, Air India was acquired by the Tata Group in 2022. The airline has expanded with hundreds of additional flights, flying 60 million customers to 103 destinations through 2024.

The new owners invested billions, and the airline has ordered hundreds of planes to replace its aging fleet.

In a December interview with BI, Wilson compared his work revitalising Air India to "drinking from a firehose."

He added that he thought the turnaround was close to completion, but said there were supply-chain constraints. "Until we upgrade the aircraft, then people won't believe that the transformation has happened," Wilson said.

Alan Tan, an aviation law professor at the National University of Singapore, told BI that Air India in particular would have an immediate hit to customer perception.

"But as other leading airlines facing crises have shown, these are not insurmountable," he added. "Transparency and accountability in investigations, and consistent messaging to the public, will hopefully reduce the risks of a media spectacle."

A lengthy investigation

It will take a thorough and lengthy investigation before there are answers about what caused the crash.

Attorneys who have battled Boeing in the courts were among the people BI spoke to who were hesitant to draw any conclusions.

"The fact that this tragedy involves a Boeing aircraft does not necessarily mean that there's something wrong with the actual aircraft — as distinguished from issues surrounding maintenance, or even products that are not Boeing's, such as the engines," said Robert Clifford, lead counsel for the families of victims of the 2019 Ethiopian Airlines crash, in which a 737 Max crashed shortly after takeoff, killing more than 150 people.

He added that a quick and efficient investigation is necessary to "help calm the public."

Thursday's incident was the first fatal crash and total hull loss of a Boeing 787 Dreamliner, one of the most advanced passenger jets, which entered service in 2011.

The model has faced some criticism from whistleblowers. Last year, Sam Salehpour, a quality engineer at Boeing, told NBC he observed "shortcuts to reduce bottlenecks" in manufacturing 787s. Boeing responded that it was "fully confident in the 787 Dreamliner."

On Thursday, Salehpour's attorneys urged the Federal Aviation Administration to release a report investigating his claims.

Richard Aboulafia, managing director at Aerodynamic Advisory, told BI, "It's a terrible tragedy, but I just don't see how this impacts anything [for Boeing]."

"Unless it's the unlikely event that they do find a design or manufacturing flaw, but after all these years, both for this type of aircraft and this particular aircraft, that's not normal," he added.

• Unwanted subscriptions were about to get easier to cancel with the FTC's new click-to-cancel rule.
• But the commission just delayed its enforcement deadline by two more months.
• Ex-FTC commissioner Lina Khan says the move lets firms "keep trapping people" in pesky subscriptions.

It was about to get easier to get rid of that pesky subscription you've been stuck paying for until the Federal Trade Commission delayed enforcement of its new click-to-cancel rule.

Former FTC chair Lina Khan, in a Thursday post on X, said that the enforcement delay will give firms more time "to keep trapping people in subscriptions."

Most consumers are familiar with the unwanted subscription rigamarole: It's painlessly simple to sign up online for a streaming service, gym, or other subscription, but when the time comes to stop monthly payments and unsubscribe, there's no way to do it digitally, and you're forced into the dreaded routine of navigating call center chatbots that only seem to operate during the middle of your workday.

The FTC's click-to-cancel rule was supposed to go into effect in its entirety this week, ending the nightmarish cycle and making it just as easy for consumers to cancel their subscriptions as it was to start them. But on Friday, the commission's leaders voted to extend its enforcement deadline by two more months.

"Having conducted a fresh assessment of the burdens that forcing compliance by this date would impose, the Commission has determined that the original deferral period insufficiently accounted for the complexity of compliance," read a statement from Chairman Andrew Ferguson, co-signed by commissioners Melissa Holyoak and Mark Meador, about the decision.

After the FTC approved the click-to-cancel rule, also known as the Negative Option Rule, in November 2024, businesses had more than six months to comply before enforcement was scheduled to begin.

The rule's requirement to remove statements that misrepresent the nature of a subscription took effect on January 14. Its enforcement provisions — requiring clear disclosures, user consent, and easy cancellation policies —  were set to take effect on May 14. However, the FTC's latest decision pushes the enforcement deadline back by 60 days, to July 14.

"We object to the delay," former FTC commissioners Alvaro Bedoya and Rebecca Slaughter said in a joint statement posted to social media on Tuesday. "And were we allowed to exercise our duties as commissioners, we would have voted 'no.'"

Bedoya and Slaughter were the only two Democrats serving as FTC commissioners until March 18, when President Donald Trump fired them. The pair, whose terminations indicated their service at the FTC was "inconsistent" with Trump's policy priorities, have filed suit against the administration, alleging their firings violate a 1935 Supreme Court precedent that the president cannot fire FTC commissioners without cause, CNN reported.

Even if Bedoya and Slaughter had remained at the FTC, the conservative majority at the commission would be able to pass rules via a 3-2 vote. The decision to delay the click-to-cancel enforcement received a 3-0 vote, with all three Republican commissioners voting in favor of the deadline extension.

"The companies create these traps," Bedoya and Slaughter's statement continued. "They're the ones who made it so hard to get out. They didn't have to wait to make it easier to unsubscribe. But they did — they waited until the FTC told them to stop. Then, they still got six months to get their houses in order. Why do they get another two months to comply?"

Representatives for the FTC did not immediately respond to a request for comment from Business Insider.