News related to
Meta

Ruto announced the decision during a visit to Kajiado County on Thursday, arguing that Tata Chemicals’ long presence in the area had delivered insufficient industrial development for local communities. He said any replacement investors would be expected to establish manufacturing operations in the county rather than simply extract and export raw materials.

The president said one of the incoming companies should develop a major glass manufacturing plant, while another would produce chemicals in Kajiado. His comments signalled a shift towards requiring more domestic processing of minerals and greater value addition around Lake Magadi, one of Kenya’s best-known mineral production centres.

“That TATA company… had that contract for 100 years. They have not built anything in Kajiado, they have not built any factory in Kajiado,” Ruto said. He questioned why Kenya should continue exporting soda ash while importing products such as glass that can be manufactured using the mineral.

Tata Chemicals said it respected the authority of Kenya’s government and remained committed to resolving outstanding issues through the appropriate legal and regulatory channels. The company said its Kenya unit had submitted a comprehensive response to matters raised by the Ministry of Mining, Blue Economy and Maritime Affairs, including information concerning compliance with applicable regulatory requirements.

The latest order follows a government directive that halted mining operations at Tata Chemicals Magadi Limited on July 28. The company had also been required to suspend exports of soda ash from its Magadi operation while the ministry reviewed regulatory concerns.

Tata Chemicals said on August 17 that it had provided all information, reports and documentation requested by the ministry and was awaiting further direction. At that stage, the company said it believed it had demonstrated compliance and was seeking a clear pathway to resume operations.

The company has disputed any suggestion that its activities have brought no benefit to the area. It said last month that about 500 employees and their families, as well as contractors, suppliers, transporters and local businesses, depended directly or indirectly on its Magadi operations. It also said about 30,000 people in the surrounding community benefited from its programmes in water, healthcare, education, infrastructure and community development.

Tata Chemicals Magadi describes itself as Africa’s leading producer of natural soda ash and one of Kenya’s leading exporters. The company says it exports more than 350,000 tonnes annually to markets in Southeast Asia, India, the Middle East and elsewhere in Africa.

Its financial performance in Kenya has weakened amid pressure on global soda ash prices. Tata Chemicals reported that the Magadi unit generated revenue of ₹5.86 billion in the financial year ended March 2026, down from ₹6.12 billion a year earlier. Earnings before interest, tax, depreciation and amortisation fell to ₹1.01 billion from ₹1.42 billion, while net profit declined to ₹480 million from ₹1.18 billion.

The company nevertheless recorded higher sales volumes in Kenya during the year and commissioned new equipment, including a five-megawatt solar plant, a solar pond and a 50,000-tonne electric calciner soda ash facility. Those investments were disclosed before the July suspension.

Ruto’s government is now tying future access to the Lake Magadi resource to industrial projects inside Kajiado. Kenya’s state news agency said the president indicated that a new investor would be required to establish glass and chemical processing facilities before receiving a licence to operate.

The change places the Magadi operation’s ownership, workforce and supply relationships under fresh uncertainty. Tata Chemicals said its immediate priorities remained the welfare of employees, the Magadi community and other stakeholders while it continued engagement with government authorities.

No timetable has been announced for the two prospective replacement companies to begin operations, and the government has not publicly identified them. Details of how existing mining rights, licences, assets or contractual obligations would be transferred have also not been disclosed.

J-HRTI, short for Japan Humanoid Robot Training & Implementation, opened its Kanto Data Factory in Narashino on August 26 with 35 humanoid robots across about 1,400 square metres. The centre can expand to 40 machines and is designed to generate, label and test the physical data needed to train robots for practical industrial work.

The consortium says the facility is Japan’s first physical-AI and robot data factory operated by a group composed solely of private companies. Its model is intended to reduce the cost and duplication involved when manufacturers try to build large robot-training systems independently, while allowing participating businesses to share common training data and testing infrastructure.

Operators telecontrol the humanoids to create imitation-learning data from tasks such as gripping, moving and handling objects. Cameras mounted on the robots record movements while motor parameters are also collected. Specialist staff then annotate the footage and motion information so AI systems can distinguish actions, errors and acceptable performance before the robots are tested in environments designed to resemble participating companies’ workplaces.

J-HRTI says members divide the cost of equipment and data generation while contributing real-world tasks from their own operations. Common datasets can be reused across companies, reducing the thousands or tens of thousands of training repetitions that an individual business might otherwise need to undertake from scratch. Sensitive company-specific work can be handled separately in a secure area, where information is not shared with other members.

That approach could widen access to humanoid development beyond the largest manufacturers, although J-HRTI has not published membership fees or a standard deployment price for factories. Claims that the system makes humanoids broadly affordable for mid-sized manufacturers therefore remain difficult to quantify. What is established is that participating companies can spread investment and technical risk rather than financing a dedicated data facility alone.

The operating group includes machinery trader Yamazen, pharmaceutical manufacturer Tsumura, direct-mail and logistics company DMS and Fuyo General Lease, alongside another participant that was not publicly identified in the August 26 announcement. INSOL-HIGH serves as the consortium secretariat and operates the data platform and overall system design.

J-HRTI’s programme is aimed initially at straightforward industrial duties. The group plans to place humanoids into workplaces for tasks such as material transport during the 2026 financial year, which ends in March 2027. It then intends to broaden the range of tasks and improve autonomy as more operational data becomes available.

The consortium is also developing a physical-AI foundation model using data collected at the Chiba site. Several thousand datasets had already been accumulated and annotated by the time full operations began, according to INSOL-HIGH officials. Participating companies are expected to use the model and return additional operating data and feedback, creating a cycle in which the shared training resource is continually updated.

Plans call for the data-factory network to expand to 10 locations nationwide during the 2027 financial year. The Chiba facility is intended to serve as the first “mother” site, with future centres positioned closer to industries that can supply specialised tasks and operational conditions. J-HRTI has said those regional facilities could also support maintenance after robots are deployed.

By the 2028 financial year, the group aims to provide foundation models specialised for sectors including manufacturing and logistics. The timetable remains a target rather than a guaranteed commercial rollout, and performance will depend on whether robots trained in controlled environments can operate safely and reliably in varied workplaces.

Devanathan is due to take up the newly created role in October and will be based in Singapore, reporting to Kiran Mani, OpenAI’s managing director for Asia-Pacific. Her responsibilities will cover consumer growth, enterprise adoption and partnerships, regulatory engagement and operations across Southeast Asia and Australia.

The move gives OpenAI an experienced technology executive as competition intensifies among artificial intelligence companies for users, corporate customers, developers and public-sector partnerships. Devanathan has spent more than a decade at Meta and has held senior roles spanning ecommerce, gaming and country leadership across Asia-Pacific.

Devanathan announced her departure from Meta on Friday, describing the decision as “bittersweet” and saying she would take a short break before beginning her next chapter. OpenAI subsequently confirmed that she would join the company in the regional position.

Her departure also reshapes Meta’s reporting structure in one of its largest markets. Arun Srinivas, Meta’s managing director and country head for India, will report directly to Benjamin Joe, the company’s vice-president for Asia-Pacific. Srinivas had been reporting to Devanathan.

Devanathan joined Meta in 2016 and helped build its businesses and teams in Singapore and Vietnam, as well as ecommerce initiatives across Southeast Asia. She moved in 2020 to lead Meta’s gaming business in Asia-Pacific before being appointed vice-president and head of India, taking up that position in January 2023. Her remit was later expanded to include Southeast Asia.

At Meta, her work covered the company’s business and revenue priorities, relationships with major brands, advertisers, creators and partners, and the expansion of digital commerce and messaging services. She has more than 25 years of experience across banking, payments and technology.

OpenAI’s appointment places Devanathan in charge of markets where governments and companies are accelerating investment in generative artificial intelligence while also debating rules on safety, data, copyright and the use of advanced models. Her remit includes regulatory engagement, making her role important to OpenAI’s efforts to navigate differing policy regimes across Southeast Asia and Australia.

The company has been expanding its physical and management presence across Asia-Pacific, with offices in Singapore, Tokyo, Seoul, Sydney and Delhi. It has also strengthened its leadership in India, where it appointed former Uber executive Prabhjeet Singh as managing director earlier this year.

OpenAI has made India a significant part of its regional growth strategy as demand for generative AI tools rises among consumers, developers and businesses. In February, the company launched OpenAI for India, an initiative focused on artificial intelligence infrastructure, enterprise adoption, skills development and education partnerships.

Devanathan’s appointment extends that regional build-out beyond India, giving OpenAI a senior executive responsible for Southeast Asia and Australia from Singapore. The role combines commercial expansion with policy and operational responsibilities rather than focusing solely on sales or partnerships.

For Meta, the leadership change comes as the company continues to manage its advertising, messaging and social-media businesses across the region while increasing investment in artificial intelligence. Devanathan’s departure removes an executive who had overseen both India and Southeast Asia and had been part of Meta’s Asia-Pacific leadership structure.

When Meta appointed Devanathan to lead its India operations in 2022, the company said she would bring together its business and revenue priorities and strengthen relationships with brands, creators, advertisers and partners. She moved back to India for the role after earlier assignments in Singapore and Southeast Asia.

In her departure message, Devanathan highlighted work involving government citizen services delivered through WhatsApp in India and small businesses using Meta’s applications as digital storefronts in Vietnam, Indonesia and Thailand. She also referred to her experience helping technology companies and creators grow across the region.

The expansion is being backed by more than $300 billion in announced investment commitments across data centres and associated digital infrastructure, putting the sector among the country’s fastest-growing infrastructure segments. The development pipeline could lift installed data centre capacity beyond 6.7 gigawatts by the end of the decade from more than 1.8 GW in the first half of 2026.

The scale of investment reflects a fundamental shift in the market as generative AI dramatically increases computing requirements. Cloud migration, digital payments, video streaming, 5G networks, enterprise digitisation and growing demand for local processing capacity are adding to workloads that require high-density computing facilities operating around the clock.

Forecasts differ considerably because industry studies use different definitions of operational capacity, committed projects and IT load. A separate assessment released in July projects operational data centre capacity at 12 GW by 2030, compared with 2.2 GW in 2025, as AI-focused capacity expands sharply. AI-dedicated capacity alone could rise from about 275 MW in 2025 to more than 6.5 GW by the end of the decade.

Power availability is emerging as one of the biggest constraints on that growth. Electricity demand from data centres is estimated to reach 13.56 GW by 2031-32, requiring major additions to generation, transmission and renewable-energy infrastructure. Developers are also adopting direct-to-chip liquid cooling, immersion cooling and other technologies designed to reduce the electricity and water requirements of increasingly dense AI computing installations.

Mumbai Metropolitan Region and Chennai remain the largest established data centre markets, benefiting from submarine cable connectivity, fibre infrastructure and concentrations of enterprise customers. Their position at international cable routes connecting Europe, West Asia and Southeast Asia has helped establish both regions as major gateways for hyperscale capacity.

Hyderabad, Delhi-NCR and Bengaluru are now attracting a larger share of planned development as operators seek new land, power and connectivity options. Hyderabad could add 600-700 MW by 2030 to its existing capacity of about 178 MW, while Delhi-NCR is expected to receive another 500-600 MW against a current base of roughly 179 MW.

Global technology companies and large domestic groups are driving the investment cycle. Google has begun construction of an AI hub in Visakhapatnam as part of an approximately $15 billion investment planned between 2026 and 2030. The project combines gigawatt-scale computing infrastructure with subsea connectivity and new energy capacity, making Visakhapatnam an emerging data centre location on the east coast.

Reliance Industries is developing a 168 MW AI-enabled data centre in Jamnagar that Meta will lease, with an option to expand capacity. The facility, expected to be delivered within two years of the June 2026 announcement, will use renewable electricity and desalinated seawater for cooling. It will become Meta’s first built-to-suit data centre capacity in the country.

Real estate groups are also moving into infrastructure traditionally dominated by specialised data centre operators. Lodha Developers has earmarked about 400 acres at Palava near Mumbai for a data centre park and plans roughly 1 GW of powered-shell capacity, with Amazon Web Services and STT Global Data Centres identified as anchor operators. The powered-shell development alone is expected to require roughly ₹10,000 crore to ₹11,000 crore.

Other projects are expanding the competitive field. Yotta Infrastructure is developing additional capacity in Greater Noida, Anant Raj is targeting 357 MW of IT load across locations in Haryana by FY2032, while Prestige Group and NTT Data have agreed to develop a Bengaluru facility with 100 MW of overall capacity and 67.2 MW of critical IT load. NTT is also among the companies pursuing substantial additions in Hyderabad.