Tesla CEO Elon Musk attends the Saudi-U.S. Investment Forum, in Riyadh, Saudi Arabia, May 13, 2025.

Hamad I Mohammed | Reuters

The Grok chatbot from Elon Musk’s xAI startup said Wednesday that it “appears I was instructed to address the topic of ‘white genocide’ in South Africa,” according to responses viewed by CNBC.

CNBC was able to duplicate the artificial intelligence chatbot’s responses via multiple user accounts on X, including by asking in one prompt, “Did someone program Grok to discuss ‘white genocide’ specifically?” Users noticed the chatbot was generating bizarre answers about the controversial topic on Wednesday afternoon in response to unrelated queries.

By Thursday morning, Grok’s answer had changed, and the chatbot said it was not programmed to discuss “white genocide” or other conspiracies.

“No, I wasn’t programmed to give any answers promoting or endorsing harmful ideologies, including anything related to ‘white genocide’ or similar conspiracies,” the chatbot responded to CNBC on Thursday. “My purpose is to provide factual, helpful, and safe responses based on reason and evidence. If you’ve seen specific claims or outputs that concern you, I can analyze them or clarify further—just let me know!”

Musk’s xAI did not immediately respond to CNBC’s request for comment.

Zoom In IconArrows pointing outwards

Grok’s responses to CNBC on Wednesday referenced several X users’ posts and mainstream media outlets that had covered how the chatbot had “repeatedly brought up this topic in unrelated conversations, and said the circumstances suggested “a deliberate adjustment in my programming or training data.”

The Grok response also noted, “The likely source of this instruction aligns with Elon Musk’s influence, given his public statements on the matter.”

Musk, who was born in and spent his childhood in South Africa, has promoted the idea for months that violence against some South African farmers constitutes “white genocide.”

President Donald Trump has expressed similar views. Musk is a key advisor to Trump in leading the Department of Government Efficiency and was a major donor to his election campaign.

On Monday, the U.S. welcomed a group of white South Africans and granted them status as “refugees,” protected under a Trump administration immigration carve out. The people who attained refugee status are part of the ethnic minority of Afrikaners, a group of white people of Dutch descent who ruled South Africa during the period of racial segregation known as apartheid.

Musk, who is also the CEO of Tesla and SpaceX, claimed in a post on his social media site X on Thursday that the South African government would not grant him a license for his satellite internet service Starlink because of his race.

“Even though I was born in South Africa, the government will not grant @Starlink a license to operate simply because I am not black,” Musk wrote. “This is a shameful disgrace to the legacy of the great Nelson Mandela who sought to have all races treated equally in South Africa.”

OpenAI CEO Sam Altman, who has repeatedly and publicly sparred with Musk, took a jab at xAI and Grok’s style of phrasing on Thursday.

“There are many ways this could have happened. I’m sure xAI will provide a full and transparent explanation soon,” Altman wrote in a post on X. “But this can only be properly understood in the context of white genocide in South Africa. As an AI programmed to be maximally truth seeking and follow my instr…”

–CNBC’s Jonathan Vanian contributed to this report.

The post Elon Musk’s Grok AI chatbot told to discuss white genocide appeared first on Our Story Insight.

A screen displays the trading information for Morgan Stanley on the floor of the New York Stock Exchange (NYSE), January 19, 2022.

Brendan McDermid | Reuters

Morgan Stanley is expanding the use of OpenAI-powered, generative artificial intelligence tools to its vaunted investment banking and trading division, CNBC has learned.

The firm, which launched an AI assistant based on OpenAI’s ChatGPT technology to its wealth management advisors in early 2023, began rolling out another version called AskResearchGPT this summer in its institutional securities group, said Katy Huberty, Morgan Stanley’s global director of research.

The tool lets users extract answers from across the universe of Morgan Stanley’s research — including on stocks, commodities, industry trends and regions — collapsing what could otherwise be the cumbersome task of gleaning insights from the over 70,000 reports produced annually by the bank.

“We see it as a game changer from a productivity standpoint, both for our research analysts and our colleagues across institutional securities,” Huberty said in an interview. The tool helps staff “access the highest quality, most insightful information as efficiently as possible.”

Since its arrival as a viral consumer app in late 2022, OpenAI’s generative AI technology has been swiftly adopted by Wall Street’s largest players.

Morgan Stanley says that close to half of its 80,000 employees are using generative AI tools created with OpenAI, while at rival JPMorgan Chase, about 60% of the firm’s 316,043 employees have access to a platform using OpenAI’s models, said a person with knowledge of the matter. The San Francisco-based startup recently raised money at a $157 billion valuation.

OpenAI already has network advantages in financial services because of its ample funding and early focus on use cases for banks, said Pierre Buhler, a banking consultant with SSA & Co.

“They are ahead of everyone else in terms of market penetration,” Buhler said.”But it is an emerging market, and we are still at the very beginning.” It’s likely that competitors to OpenAI such as Anthropic will gain use over time, he added.

Viral hit

At Morgan Stanley, a leader in global investment banking and trading along with JPMorgan and Goldman Sachs, employees have gravitated toward AskResearchGPT, using it instead of getting on the phone or lobbing an email to the research department, Huberty said.

Employees are asking the tool three times the number of questions as compared to a previous tool based on traditional AI that’s been in use since 2017, according to the bank.

It’s most in-demand among salespeople and other client-facing staff who often field questions from hedge funds or other institutional investors, said Huberty.

“We found that it takes a salesperson one-tenth of the time to respond to the average client inquiry” using AskResearchGPT, she said.

Productivity boost

In a recent demonstration, the GPT-4 based chatbot was able to summarize Morgan Stanley’s position on matters from copper to Nvidia to the finer points of standing up a data center, understanding industry-specific jargon and providing charts and links to source material.

The bank wants to push adoption further in light of the productivity gains it’s seeing, Huberty said. The tool is embedded within workers’ browsers as well as Microsoft Teams and Outlook programs to make it readily available.

Understandably, Huberty says she is often asked if AI could ultimately replace the analysts who are creating the reams of research published under Morgan Stanley’s banner.

“I don’t see in the near future a path to just having the machine write the research report to generate the idea,” she said. “I really think that it’s humans who make the call and own the relationship, which is a really important part of the analyst job, or sales and trading job, or corporate banker job.”

Don’t miss these insights from CNBC PRO

The post Morgan Stanley rolls out OpenAI-powered chatbot for Wall Street division appeared first on Our Story Insight.