The logo of telecoms giant Orange displayed at Mobile World Congress 2024 in Barcelona, Spain.
Joan Cros | Nurphoto via Getty Images
French telecoms giant Orange on Tuesday said it’s partnering with Microsoft-backed OpenAI and Facebook-owner Meta to build custom artificial intelligence models designed to better understand regional African languages.
Orange said it’s working with OpenAI and Meta to develop custom AI models built on their respective Whisper and Llama open-source AI models — openly available systems that can be adapted to meet specific needs — that can understand West African languages not understood by most conversational systems.
Currently, much of the data major AI companies train their algorithms on originates in the United States, which means their models can lose important context, such as culture and language, when it comes to different regions like Europe, the Middle East and Africa.
That means it can be hard for those models to understand text and voice-based communications composed in less well-represented languages, according to Steve Jarrett, Orange’s chief AI officer.
“Having an open model, you’re able to do what’s called fine tuning, where you you introduce additional information to the model that wasn’t included when it was first trained,” Jarrett told CNBC in an interview. “We’re adding the recognition of West African regional languages that are not understood today by any AI.”
Orange plans to start by rolling out AI models that incorporate two West African regional languages, Wolof and Pulaar, which are spoken by roughly 16 million people and six million people, respectively, in early 2025.
Wolof is a language spoken in Senegal, the Gambia and southern Mauritania, while Pulaar is mostly spoken in Senegal.
The open-source AI models will be provided externally by Orange with a free license for non-commercial uses including public health and education, the company said. Orange plans to expand its custom AI model initiative to eventually cover all 18 West African countries.
“We’re operating in West African countries where a lot of these regional languages are being spoken in our contact centers, but where the current AI models don’t understand what these people are typing or saying,” Jarrett told CNBC.
Major large language models like OpenAI’s GPT, Meta’s Llama and Anthropic’s Claude aren’t well suited to Africans’ needs as they weren’t trained specifically on data originating from the region, according to Orange’s AI chief.
‘Sovereign AI’ push
The move taps into a concept that’s been gaining traction globally, known as “sovereign AI.”
The term refers to the idea that individual countries and regions should seek greater control over the core technological infrastructure upon which AI systems are built, by localizing data storage and processing to ensure they represent specific languages, culture and history.
Orange is also looking to localize data processing and the hosting of OpenAI’s models in European data centers. This, Orange said, will give it early access to OpenAI’s latest and most advanced AI models and help it build new applications such as AI-powered voice systems for customer service.
Jarrett said Orange is committing to using AI “responsibly” and “not always using the massive, large language model [LLM] for every problem” given environmental concerns associated with the technology’s huge energy requirements.
In addition to using AI systems to improve customer service, Orange is also using the tech to improve a core part of its business: mobile networks.
“On the network side, we use [AI] to not only optimize how we plan the network, but also how we operate the network right,” Jarrett told CNBC.
“The volume of data is so large coming from all the network equipment that with AI systems, we can help identify those patterns in the data that could help us identify and predict failures even before the customer notices.”
Google announced Monday the removal of nearly 11,000 YouTube channels and other accounts tied to state-linked propaganda campaigns from China, Russia and more in the second quarter.
The takedown included more than 7,700 YouTube channels linked to China.
These campaigns primarily shared content in Chinese and English that promoted the People’s Republic of China, supported President Xi Jinping and commented on U.S. foreign affairs.
Over 2,000 removed channels were linked to Russia. The content was in multiple languages that supported Russia and criticized Ukraine, NATO and the West.
Google, in May, removed 20 YouTube channels, 4 Ads accounts, and 1 Blogger blog linked to RT, the Russian state-controlled media outlet accused of paying prominent conservative influencers for social media content ahead of the 2024 election.
Tim Pool, Dave Rubin and Benny Johnson — all staunch supporters of President Donald Trump — made content for Tenent Media, the Tennessee company described in the indictment, according to NBC News.
Read more CNBC tech news
YouTube began blocking RT channels in March 2022, shortly after Russia invaded Ukraine.
The active removal of accounts is part of the Google Threat Analysis Group’s work to counter global disinformation campaigns and “coordinated influence” operations.
Google’s second quarter report also outlined the removal of influence campaigns linked to Azerbaijan, Iran, Turkey, Israel, Romania and Ghana that were found to be targeting political rivals.
Some campaigns centered on growing geopolitical conflicts, including narratives on both sides of the Israel-Palestine War.
CNBC has reached out to YouTube for further comment or information on the report.
Google took down more than 23,000 accounts in the first quarter.
Meta announced last week it removed about 10 million profiles for impersonating large content producers through the first half of 2025 as part of an effort by the company to combat “spammy content.”
Chris Martin of Coldplay performs live at San Siro Stadium, Milan, Italy, in July 2017.
Mairo Cinquetti | NurPhoto | Getty Images
Astronomer‘s interim CEO said in his first public comment since unexpectedly taking over the role on Saturday that he hopes to move the tech startup past the viral moment that captured national attention last week.
Pete DeJoy was appointed to the top job due to the resignation of CEO Andy Byron, days after he was caught on video in an intimate moment with the company’s head of human resources at a Coldplay concert. Astronomer said over the weekend that it would begin a search for a new CEO.
“The events of the past few days have received a level of media attention that few companies — let alone startups in our small corner of the data and AI world — ever encounter,” DeJoy wrote in a LinkedIn post on Monday. “The spotlight has been unusual and surreal for our team and, while I would never have wished for it to happen like this, Astronomer is now a household name.”
Byron was shown on a big screen at the concert in Boston on Wednesday with his arms around Chief People Officer Kristin Cabot. Byron, who is married with children, immediately hid when the couple was shown on screen. Lead singer Chris Martin said, “Either they’re having an affair or they’re just very shy.” A concert attendee’s video of the affair went viral.
Read more CNBC reporting on AI
DeJoy helped start Astronomer in 2017, according to his LinkedIn profile, and had been serving as chief product officer since earlier this year.
In May, Astronomer announced a $93 million investment round led by Bain Ventures and other investors, including Salesforce Ventures.
“I’m stepping into this role with a wholehearted commitment to taking care of our people and delivering for our customers,” DeJoy wrote. He added that “our story is very much still being written.”
Astronomer is commercializing the open-source data operations platform Astro. DeJoy wrote that customers “trust us with their most ambitious data & AI projects” and that “we’re here because the mission is bigger than any one moment.”
Dylan Field, co-founder and CEO of Figma Inc., after the morning sessions at the Allen & Co. Media and Technology Conference in Sun Valley, Idaho, on July 11, 2024.
David Paul Morris | Bloomberg | Getty Images
Design software company Figma on Monday published an updated prospectus for its initial public offering.
The company said it expects to sell about 37 million shares at $25 to $28 each. That would generate as much as $1 billion in proceeds, between the company and selling shareholders.
The IPO could value Figma, led by co-founder Dylan Field, a fully diluted valuation of $14.6 billion to $16.4 billion. Field plans to sell 2.35 million shares, which could be worth as much as $65.8 million.
In a 2024 tender offer, investors valued the company at $12.5 billion. In 2022, Adobe had agreed to acquire Figma for $20 billion, but the deal was scrapped after regulators objected.
The flow of technology companies joining U.S. exchanges has slowed since late 2021. Concerns over inflation and a recession made some investors less interested in backing fast-growing but money-losing companies.
But a few technology stocks have become available in recent months. CoreWeave went public in March, and Circle and Chime shares started trading in June.
Read more CNBC tech news
Figma filed to go public on July 1, announcing plans to trade on the New York Stock Exchange under the symbol “FIG.”
On Monday, it provided preliminary results for the second quarter, showing $9.0 million to $12.0 million in operating income on $247 million to $250 million in revenue. That would imply year-over-year revenue growth of 39% at the low end and 41% at the high end. Growth in the first quarter exceeded 46%.
During the second quarter, Figma added clients and expanded business with existing ones. The company’s operating margin would be ticking up to 4% to 5%, up from 3% in the same quarter a year ago, based on the preliminary results.
Figma said it has authorized the issuance of “blockchain common stock” in the form of “blockchain-based tokens.” So far, though, Figma said it isn’t planning to issue this type of stock. In July, Figma disclosed investments in a stablecoin and a Bitcoin exchange-traded fund.
Mike Krieger, a co-founder of Instagram who is now chief product officer of artificial intelligence model developer Anthropic, has joined the board. Luis von Ahn, co-founder and CEO of Duolingo, is also joining the board, according to the filing.