Digital cooperation

Google invites African AI startups to join 2025 accelerator

Google has opened applications for its 2025 Startups Accelerator Africa programme, aiming to support early-stage African companies leveraging AI to solve critical local challenges.

The three-month initiative is open to Seed to Series A startups with a live product, at least one founder of African descent, and a strong focus on responsible AI development.

Successful applicants will receive tailored technical mentorship, up to $350,000 in Google Cloud credits, and access to a global network of investors and partners.

Participants will also benefit from workshops covering technology, strategy, leadership, and AI implementation.

Google emphasised that empowering local innovators is key to unlocking the potential of AI across the continent.

Since 2018, the programme has backed 140 startups from 17 countries, helping them raise over $300 million and create more than 3,000 jobs.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Chinese AI sector eyes Huawei’s powerful 910C chip

Huawei is preparing to begin large-scale shipments of its new AI chip, the 910C, as early as next month.

The move is seen as a critical development in China’s ongoing effort to reduce reliance on US technology, especially after recent export restrictions on Nvidia’s H20 chip.

The 910C delivers performance on par with Nvidia’s H100 by combining two of Huawei’s earlier 910B chips in a single integrated package.

These enhancements, which include increased computing power and memory capacity, have attracted interest from Chinese AI developers scrambling for alternatives.

Despite Huawei’s refusal to confirm the reports, sources say sample shipments have already taken place. Some chips are being produced domestically using SMIC’s 7nm process, while others may involve components originally manufactured by TSMC, raising fresh regulatory scrutiny.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

AI expert Aidan Gomez joins Rivian board

Aidan Gomez, co‑founder and chief executive of AI specialist Cohere, has been appointed to the board of electric‑vehicle maker Rivian, according to a recent regulatory filing.

Rivian expanded its board and elected Gomez for a term running until 2026, signalling the company’s intent to deepen its software credentials.

Gomez brings a distinguished AI pedigree, having co‑authored the seminal 2017 paper ‘Attention Is All You Need’ and led research at Google Brain before launching Cohere in 2019.

Under his leadership, Cohere has trained large‐scale foundation models for enterprise clients such as Oracle and Notion, positioning it at the forefront of generative AI.

Rivian is already collaborating on a $5.8 billion joint venture with Volkswagen to develop and licence its electrical architecture and software. Parallel efforts include the creation of an in‑vehicle AI assistant, overseen by Rivian’s chief software officer, Wassym Bensaid.

Founder and CEO RJ Scaringe praised Gomez’s expertise as vital for integrating ‘cutting‑edge technologies into our products, services, and manufacturing.’

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

AI startup caught in Dev Mode trademark row

Figma has issued a cease-and-desist letter to Swedish AI startup Loveable over the use of the term ‘Dev Mode,’ a name Figma trademarked in 2023.

Loveable recently introduced its own Dev Mode feature, prompting the design platform to demand the startup stop using the name, citing its established use and intellectual property rights.

Figma’s version of Dev Mode helps bridge the gap between designers and developers, while Loveable’s tool allows users to preview and edit code without linking to GitHub.

Despite their differing functions, Figma insists on protecting the trademark, even though ‘developer mode’ is a widely used phrase across many software platforms. Companies such as Atlassian and Wix have used similar terminology long before Figma obtained the trademark.

The legal move arrives as Figma prepares for an initial public offering, following Adobe’s failed acquisition attempt in 2023. The sudden emphasis on brand protection suggests the company is taking extra care with its intellectual assets ahead of its potential stock market debut.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

CISA extends MITRE’s CVE program for 11 months

The US Cybersecurity and Infrastructure Security Agency (CISA) has extended its contract with the MITRE Corporation to continue operating the Common Vulnerabilities and Exposures (CVE) program for an additional 11 months. The decision was made one day before the existing contract was set to expire.

A CISA spokesperson confirmed that the agency exercised the option period in its $57.8 million contract with MITRE to prevent a lapse in CVE services. The contract, which originally concluded on April 17, includes provisions for optional extensions through March 2026.

‘The CVE Program is invaluable to the cyber community and a priority of CISA,’ the spokesperson stated, expressing appreciation for stakeholder support.

Yosry Barsoum, vice president of MITRE and director of its Center for Securing the Homeland, said that CISA identified incremental funding to maintain operations.

He noted that MITRE remains committed to supporting both the CVE and CWE (Common Weakness Enumeration) programs, and acknowledged the widespread support from government, industry, and the broader cybersecurity community.

The extension follows public concern raised earlier this week after Barsoum issued a letter indicating that program funding was at risk of expiring without renewal.

MITRE officials noted that, in the event of a contract lapse, the CVE program website would eventually go offline and no new CVEs would be published. Historical data would remain accessible via GitHub.

Launched in 1999, the CVE program serves as a central catalogue for publicly disclosed cybersecurity vulnerabilities. It is widely used by governments, private sector organisations, and critical infrastructure operators for vulnerability identification and coordination.

Amid recent uncertainty about the program’s future, a group of CVE Board members announced the formation of a new non-profit organisation — the CVE Foundation — aimed at supporting the long-term sustainability and governance of the initiative.

In a public statement, the group noted that while US government sponsorship had enabled the program’s growth, it also introduced concerns around reliance on a single national sponsor for what is considered a global public good.

The CVE Foundation is intended to provide a neutral, independent structure to ensure continuity and community oversight.

The foundation aims to enhance global governance, eliminate single points of failure in vulnerability management, and reinforce the CVE program’s role as a trusted and collaborative resource. Further information about the foundation’s structure and plans is expected to be released in the coming days.

CISA did not comment on the creation of the CVE Foundation. A MITRE spokesperson indicated the organisation intends to work with federal agencies, the CVE Board, and the cybersecurity community on options for ongoing support.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

AI firm DeepSeek opens up on model deployment tech

Chinese AI startup DeepSeek has announced its intention to share the technology behind its internal inference engine, a move aimed at enhancing collaboration within the open-source AI community.

The company’s inference engine and training framework have played a vital role in accelerating the performance and deployment of its models, including DeepSeek-V3 and R1.

Built on PyTorch, DeepSeek’s training framework is complemented by a modified version of the vLLM inference engine originally developed in the US at UC Berkeley.

While the company will not release the full source code of its engine, it will contribute its design improvements and select components as standalone libraries.

These efforts form part of DeepSeek’s broader open-source initiative, which began earlier this year with the partial release of its AI model code.

Despite this contribution, DeepSeek’s models fall short of the Open Source Initiative’s standards, as the training data and full framework remain restricted.

The company cited limited resources and infrastructure constraints as reasons for not making the engine entirely open-source. Still, the move has been welcomed as a meaningful gesture towards transparency and knowledge-sharing in the AI sector.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Nvidia brings AI supercomputer production to the US

Nvidia is shifting its AI supercomputer manufacturing operations to the United States for the first time, instead of relying on a globally dispersed supply chain.

In partnership with industry giants such as TSMC, Foxconn, and Wistron, the company is establishing large-scale facilities to produce its advanced Blackwell chips in Arizona and complete supercomputers in Texas. Production is expected to reach full scale within 12 to 15 months.

Over a million square feet of manufacturing space has been commissioned, with key roles also played by packaging and testing firms Amkor and SPIL.

The move reflects Nvidia’s ambition to create up to half a trillion dollars in AI infrastructure within the next four years, while boosting supply chain resilience and growing its US-based operations instead of expanding solely abroad.

These AI supercomputers are designed to power new, highly specialised data centres known as ‘AI factories,’ capable of handling vast AI workloads.

Nvidia’s investment is expected to support the construction of dozens of such facilities, generating hundreds of thousands of jobs and securing long-term economic value.

To enhance efficiency, Nvidia will apply its own AI, robotics, and simulation tools across these projects, using Omniverse to model factory operations virtually and Isaac GR00T to develop robots that automate production.

According to CEO Jensen Huang, bringing manufacturing home strengthens supply chains and better positions the company to meet the surging global demand for AI computing power.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Zhipu AI launches free agent to rival DeepSeek

Chinese AI startup Zhipu AI has introduced a free AI agent, AutoGLM Rumination, aimed at assisting users with tasks such as web browsing, travel planning, and drafting research reports.

The product was unveiled by CEO Zhang Peng at an event in Beijing, where he highlighted the agent’s use of the company’s proprietary models—GLM-Z1-Air for reasoning and GLM-4-Air-0414 as the foundation.

According to Zhipu, the new GLM-Z1-Air model outperforms DeepSeek’s R1 in both speed and resource efficiency. The launch reflects growing momentum in China’s AI sector, where companies are increasingly focusing on cost-effective solutions to meet rising demand.

AutoGLM Rumination stands out in a competitive landscape by being freely accessible through Zhipu’s official website and mobile app, unlike rival offerings such as Manus’ subscription-only AI agent. The company positions this move as part of a broader strategy to expand access and adoption.

Founded in 2019 as a spinoff from Tsinghua University, Zhipu has developed the GLM model series and claims its GLM4 has surpassed OpenAI’s GPT-4 on several evaluation benchmarks.

In March, Zhipu secured major government-backed investment, including a 300 million yuan (US$41.5 million) contribution from Chengdu.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

US exempts key electronics from China import taxes

Smartphones, computers, and key tech components have been granted exemption from the latest round of US tariffs, providing relief to American technology firms heavily reliant on Chinese manufacturing.

The decision, which includes products such as semiconductors, solar cells, and memory cards, marks the first major rollback in President Donald Trump’s trade war with China.

The exemptions, retroactively effective from 5 April, come amid concerns from US tech giants that consumer prices would soar.

Analysts say this move could be a turning point, especially for companies like Apple and Nvidia, which source most of their hardware from China. Industry reaction has been overwhelmingly positive, with suggestions that the policy shift could reshape global tech supply chains.

Despite easing tariffs on electronics, Trump has maintained a strict stance on Chinese trade, citing national security and economic independence.

The White House claims the reprieve gives firms time to shift manufacturing to the US. However, electronic goods will still face a separate 20% tariff due to China’s ties to fentanyl-related trade. Meanwhile, Trump insists high tariffs are essential leverage to renegotiate fairer global trade terms.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Google rolls out AI to improve US power grid connections

Google has announced a partnership with PJM Interconnection, the largest electricity grid operator in North America, to deploy AI aimed at reducing delays in connecting new power sources to the grid. The move comes as energy demand surges due to the expansion of data centres required for AI development.

Wait times for connecting renewable and traditional energy sources, such as wind, solar and gas, have reached record levels, increasing the risk of blackouts and rising energy costs in the US. Google’s AI technology, developed alongside Alphabet-backed Tapestry, will streamline and automate key planning processes traditionally handled manually by grid operators.

Initial deployment will focus on automating tasks like assessing project viability, which are currently time-consuming. Over time, the project aims to create a digital model of PJM’s grid, similar to Google Maps, allowing planners to view layered data and make faster, more informed decisions.

While it is too early to quantify exactly how much time will be saved, the collaboration is expected to gradually improve planning efficiency. PJM’s grid serves 67 million people, including the world’s largest data centre hub in northern Virginia, making this a critical step toward modernising the energy infrastructure needed to support the AI era.

For more information on these topics, visit diplomacy.edu.