Orange Partners with OpenAI and Meta to Develop AI Models for African Languages

Orange Partners With OpenAI and Meta to Expand AI Models for African Regional Languages

Open-source AI models to support non-commercial uses like education and health. French telecommunications company Orange, serving 292 million customers worldwide as of September 30, 2024, announced on November 26 its partnership with OpenAI and Meta to fine-tune AI large language models (LLMs) for understanding African regional languages that are currently unsupported by any generative AI model. Orange aims to expand open-source AI models to include African regional languages for the first time, promoting digital inclusion. Also Read: Orange Business Launches GenAI Solution to Simplify AI Adoption for Businesses Orange plans to fine-tune OpenAI's open-source speech models and Meta's openly available Llama 3.1 model to support regional African languages. This project aims to develop custom AI models that enable customers to communicate naturally in their local languages with Orange for customer support and sales. These AI models will also be made available under a free license for non-commercial uses, such as public health, education, and other community services, Orange confirmed. Also Read: Nvidia Unveils New AI Model Fugatto That Generates Audio from Text and Audio Orange intends to help drive AI innovation in these regional languages including by collaborating on these new AI models with local startups and other technology companies, and by doing so, to mitigate the growing digital divide faced by people all across the African continent. The project is set to launch in the first half of 2025, with an initial focus on Wolof and Pulaar, spoken by approximately 16 million and six million people, respectively, in West Africa. Orange said its long-term goal is to collaborate with AI providers to develop models that recognise all African languages spoken and written across its 18-country footprint in the region. Also Read: Deutsche Telekom Turns Anonymised Network Data into Music with AI "By fine-tuning leading AI models such as OpenAI's 'Whisper' speech model and Meta's 'Llama' text model with diverse examples of these languages, we will enable them to better understand these regional languages. Orange's vision is to make AI and other related advances accessible to all, including illiterate populations, who are currently unable to benefit from the potential of artificial intelligence. The initiative is a blueprint for how AI can be used to benefit those currently excluded," Orange said. In addition to this regional African language recognition project, OpenAI and Orange have signed an agreement that will provide Orange with direct access to OpenAI's models, available for the first time in Europe with data processing and hosting in European data centres, enabling Orange to work on improving existing solutions across its footprint. Furthermore, Orange said this new partnership will also facilitate early access to OpenAI's latest and most advanced AI models, enabling the realisation of other key use cases such as AI-based voice interactions with Orange customers. Also Read: Telefonica's Wayra Invests in AI Startup Perplexity, Signs Commercial Agreement Orange reiterated its commitment to 'Responsible AI,' where the company carefully chooses the most appropriate and simplest solution for each AI use case. Orange concluded by saying it is committed to playing a key role in the development of Responsible AI in Africa as well as promoting open-source AI in Europe.

Orange partners with OpenAI, Meta to develop custom African-language AI models

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 enlists Meta and OpenAI to develop AI language models in Africa

(Reuters) - Orange will enlist OpenAI and Meta to fine-tune AI Large Language Models (LLMs) to translate regional African languages for the French telecoms operator, it said on Tuesday. The project is slated to start in the first half of 2025 and will initially focus on incorporating regional languages spoken in West Africa into OpenAI's "Whisper" and Meta's "Llama" software, Orange said. (Reporting by Gianluca Lo Nostro in Gdansk; Editing by David Goodman)

Orange Partners With OpenAI to Develop New AI Use Cases | PYMNTS.com

The agreement will provide Orange with direct access that includes data processing and hosting in European data centers and will facilitate early access to OpenAI's latest AI models, Orange said in a Tuesday (Nov. 26) press release. With this partnership and access, Orange will work to improve its existing solutions and develop new ones for other use cases, according to the release. For example, the firm will explore AI-based voice interactions with its customers. "Orange is focused on delivering 'Responsible AI,' where the company carefully chooses the most appropriate and simplest solution for each AI use case," the release said. "This approach means only using the latest Large Language Models where they are necessary and otherwise choosing simpler and cheaper solutions, thereby minimizing the impact on the environment as well as reducing cost for the many valuable AI use cases deployed across Orange." This announcement comes at a time when enterprise spending on generative AI has increased sixfold over the past year to reach $13.8 billion. When reporting this data on Nov. 20, venture capital firm Menlo Ventures said 72% of enterprise IT decision-makers from companies with 50 or more employees expect to see broader adoption of generative AI "in the near term." Menlo Ventures Partner Joff Redfern said at the time in a press release that organizations are "moving beyond pilots to embedding AI at the core of their business strategies." Orange also said in its Tuesday press release that it will partner with OpenAI and Meta to fine-tune large language models (LLMs) to understand regional languages in Africa, starting with the Wolof and Pulaar languages that are spoken by 16 million people and 6 million people, respectively. The telecommunications firm will use these AI models to deliver customer support and sales messages and will provide these models for use in public health, public education and other services, according to the release. "Orange's long-term goal is to work with many AI technology providers to enable future models to recognize all African languages spoken and written across Orange's 18-country footprint in the region," the release said.

Orange signs deal with OpenAI to get access to pre-release AI models

(Reuters) - Orange has struck a multi-year partnership with OpenAI in Europe that will give the French telecoms operator access to pre-release AI models, group chief artificial intelligence officer Steve Jarrett said on Wednesday. WHY IT'S IMPORTANT Orange will become the first telecoms firm in Europe to have direct access to OpenAI's models. KEY QUOTES "OpenAI's models are the most popular. And so it made financial sense for us to have a direct billing relationship", Jarrett told Reuters in an interview. "We have the ability to have access to pre-release versions of their models. We have the ability to influence the road map .... Those models are all served from secure infrastructure that's hosted in Europe," he said. He added that over 50,000 Orange employees currently use OpenAI models. CONTEXT Orange announced on Tuesday it had signed an agreement with Meta and OpenAI to translate regional African languages for the telecoms group. The provider will share data samples in Wolof and Pular to train Llama and Whisper, respectively Meta and OpenAI large language models (LLM). Orange will use the models to include these languages in its customer support, and will also outsource them to non-commercial clients such as governments, universities and startups. (Reporting by Gianluca Lo Nostro in Gdansk. Editing by Mark Potter)

OpenAI, Meta and Orange to Train AI Models on African Languages

OpenAI and Meta Platforms Inc. will begin training artificial intelligence programs on African languages, addressing a shortage of models for the continent's thousands of dialects. The project, which also involves French telecommunications operator Orange SA, will start during the first half of next year, the companies said in a statement on Tuesday. Initially, the companies will focus on two West African languages, Wolof and Pulaar, which are spoken by 22 million people in the region.

Meta, OpenAI, Orange Join Forces To Build AI Models For African Languages - Meta Platforms (NASDAQ:META), IBM (NYSE:IBM)

Orange's initiative will begin by including two West African languages, Wolof and Pulaar. French telecom company Orange announced its partnership with OpenAI, backed by Microsoft Corp MSFT and Meta Platforms Inc META, Facebook's parent company. This collaboration aims to create custom artificial intelligence (AI) models tailored to understand African regional languages better, starting with West Africa. These AI systems will be built upon open-source models, Whisper and Llama, to address the challenges of less represented languages in mainstream AI solutions. Also Read: Analog Devices Q4 Earnings: Beats Estimates, Automotive Rebound, Cautiously Optimistic Outlook & More The project highlights the difficulties major AI firms face when their systems rely predominantly on U.S.-based data, which limits their ability to interpret languages and contexts from regions like Africa, the Middle East, and Europe. Orange plans to fine-tune these models by integrating language-specific data, improving their capacity to recognize and process regional dialects, CNBC reports. Orange's initiative will begin by including two West African languages, Wolof and Pulaar, spoken by approximately 16 million and 6 million people, respectively. These languages are commonly used in Senegal, The Gambia, and Mauritania. The AI models will initially roll out in early 2025, with plans to cover all 18 West African nations in future expansions. Orange will also make these models available for non-commercial uses, including public health and education. The initiative aligns with the growing global emphasis on "sovereign AI," which focuses on creating regionally controlled AI systems to safeguard local languages, cultures, and data. Orange is also working on hosting OpenAI models in European data centers to ensure localized data processing and early access to advanced AI models. These efforts will help Orange develop innovative applications like AI-powered voice systems for customer service. Nvidia chief Jensen Huang expressed solidarity with sovereign AI, which he deems critical for countries to leverage their strengths. He is backing it with a $110 million fund for startups. International Business Machines Corp IBM chief Arvind Krishna and Nvidia CFO Colette Kressexpressed similar opinions. Also Read: Taiwan Semi's Growth Fuels US Rise To Second In Chipmaking: Report Image: Pixabay Market News and Data brought to you by Benzinga APIs