Access

US tightens AI chip export rules to maintain edge over China

The US government has announced new restrictions on exporting AI chips and technology, seeking to safeguard its dominance in AI development while limiting China’s access to advanced computing capabilities. The regulations, unveiled during the final days of President Biden’s administration, impose strict caps on AI chip exports to most countries, with exemptions for close allies such as Japan, the UK, and South Korea. Countries like China, Russia, Iran, and North Korea remain barred from accessing this critical technology.

Commerce Secretary Gina Raimondo emphasised the importance of maintaining US leadership in AI to support national security and economic interests. The regulations, which build on a four-year effort to block China’s acquisition of advanced chips, also close existing loopholes and enforce tighter controls. New limits target advanced graphics processing units (GPUs), essential for training AI models, and introduce worldwide licensing requirements for cutting-edge AI technologies. Major cloud providers like Microsoft and Amazon will face new authorisation processes to establish data centres globally under stringent conditions.

Industry leaders, including Nvidia, have expressed concerns over the broad scope of the rules, warning of potential harm to innovation and market dynamics. Nvidia called the restrictions an “overreach,” while Oracle cautioned that the measures could inadvertently benefit Chinese competitors. Despite this criticism, US officials argue the rules are vital for maintaining a competitive edge, given AI’s transformative potential in sectors like healthcare, cybersecurity, and defence. China’s Commerce Ministry condemned the move, vowing to protect its interests in response to the escalating technology standoff.

Lebanese engineer creates AI chatbot to aid displaced families

As war forced thousands of Lebanese families to flee their homes, mechanical engineer Hania Zataari developed an AI chatbot to streamline aid distribution. The tool, linked to WhatsApp, collects requests for essentials like food, blankets, and medicine, helping volunteers reach those in need more efficiently. With support from donors abroad, the project has delivered hundreds of aid packages to displaced families in Sidon and beyond.

Many displaced people have struggled to access government assistance, leaving volunteers to fill the gap. Economic turmoil has further strained resources, with aid organisations warning of severe funding shortages. Despite these challenges, the chatbot has helped distribute crucial supplies, with volunteers working tirelessly to match demand with available resources.

Researchers see potential in the technology but question its scalability in other regions. The chatbot’s success, they argue, lies in its local adaptation and cultural familiarity. While it cannot solve Lebanon’s crisis, for the families relying on it, the tool has made survival a little easier.

Father of Molly Russell urges UK to strengthen online safety laws

Ian Russell, father of Molly Russell, has called on the UK government to take stronger action on online safety, warning that delays in regulation are putting children at risk. In a letter to Prime Minister Sir Keir Starmer, he criticised Ofcom’s approach to enforcing the Online Safety Act, describing it as a “disaster.” Russell accused tech firms, including Meta and X, of prioritising profits over safety and moving towards a more dangerous, unregulated online environment.

Campaigners argue that Ofcom’s guidelines contain major loopholes, particularly in addressing harmful content such as live-streamed material that promotes self-harm and suicide. While the government insists that tech companies must act responsibly, the slow progress of new regulations has raised concerns. Ministers acknowledge that additional legislation may be required as AI technology evolves, introducing new risks that could further undermine online safety.

Russell has been a prominent campaigner for stricter online regulations since his daughter’s death in 2017. Despite the Online Safety Act granting Ofcom the power to fine tech firms, critics believe enforcement remains weak. With concerns growing over the effectiveness of current safeguards, pressure is mounting on the government to act decisively and ensure platforms take greater responsibility in protecting children from harmful content.

Italy weighs Starlink deal for secure satellite communications

Italy is exploring a potential agreement with Elon Musk’s Starlink to provide secure satellite communications for government and defence officials. The proposed five-year deal, worth €1.5 billion, would enable encrypted communications in high-risk areas. An Italian representative for Musk’s aerospace businesses stated that the country would retain full control over its data while using the technology.

Opposition parties have criticised the project, questioning whether a company owned by the US billionaire should handle sensitive Italian communications. Prime Minister Giorgia Meloni, who has faced scrutiny over her ties with Musk, dismissed concerns, insisting that discussions about the deal began before her government took office in 2022.

Starlink, a subsidiary of SpaceX, operates 6,700 satellites in low-Earth orbit and is a dominant player in global satellite communications. While the Italian government evaluates the deal, officials maintain that national interests and data security would remain protected.

Lemon8 gains popularity amid TikTok uncertainty

As the possibility of a US TikTok ban looms, social media influencers are increasingly turning to Lemon8, a new app owned by TikTok’s parent company, ByteDance, as a potential alternative. Lemon8, which launched in the US and UK in 2023, combines the best aspects of Instagram and Pinterest, offering a “lifestyle community” with an emphasis on aesthetically pleasing images, videos, and lifestyle topics like beauty, fashion, food, travel, and pets. With over 1 million daily active users in the US, it has quickly gained traction, especially among Gen Z users.

Influencers are particularly drawn to Lemon8’s integration with TikTok, allowing creators to easily cross-post and boost engagement. Despite the platform’s appeal, however, Lemon8’s future remains uncertain. Like TikTok, it is owned by ByteDance, making it potentially subject to the same US regulations, including a law requiring the company to divest from TikTok or face a ban. This uncertainty is causing anxiety among creators who fear the loss of their primary platform and are seeking safer options like Lemon8.

The app itself is gaining attention for its simplicity and visual appeal. Lemon8 stands out by offering a quieter, less chaotic environment compared to the bustling, ad-heavy content on Instagram and TikTok. Its user interface is designed for easy scrolling, and the app encourages creativity through tools that enhance text, stickers, and music, making posts feel inspirational. While it’s still early days, Lemon8 offers a nostalgic, aesthetically curated space for users who may be growing weary of the larger social media giants.

Though the app is still new, it could provide a refreshing change from the current social media landscape, where content can often feel oversaturated or too commercialised. For now, Lemon8 offers a simpler, more intentional way to engage with online content—a return to a more “authentic” era of social media, reminiscent of earlier Instagram days. Whether it will succeed in the long term remains to be seen, but for now, it’s carving out a niche for users seeking a quieter digital space.

Meta accused of using pirated books for AI

A group of authors, including Ta-Nehisi Coates and Sarah Silverman, has accused Meta Platforms of using pirated books to train its AI systems with CEO Mark Zuckerberg’s approval. Newly disclosed court documents filed in California allege that Meta knowingly relied on the LibGen dataset, which contains millions of pirated works, to develop its large language model, Llama.

The lawsuit, initially filed in 2023, claims Meta infringed on copyright by using the authors’ works without permission. The authors argue that internal Meta communications reveal concerns within the company about the dataset’s legality, which were ultimately overruled. Meta has not yet responded to the latest allegations.

The case is one of several challenging the use of copyrighted materials to train AI systems. While defendants in similar lawsuits have cited fair use, the authors contend that newly uncovered evidence strengthens their claims. They have requested permission to file an updated complaint, adding computer fraud allegations and revisiting dismissed claims related to copyright management information.

US District Judge Vince Chhabria has allowed the authors to file an amended complaint but expressed doubts about the validity of some new claims. The outcome of the case could have broader implications for how AI companies utilise copyrighted content in training data.

Brazil’s Lula criticises Meta’s move to end US fact-checking program

Brazilian President Luiz Inácio Lula da Silva has condemned Meta’s decision to discontinue its fact-checking program in the United States, calling it a grave issue. Speaking in Brasília on Thursday, Lula emphasised the need for accountability in digital communication, equating its responsibilities to those of traditional media. He announced plans to meet with government officials to discuss the matter.

Meta’s recent decision has prompted Brazilian prosecutors to seek clarification on whether the changes will affect the country. The company has been given 30 days to respond as part of an ongoing investigation into how social media platforms address misinformation and online violence in Brazil.

Justice Alexandre de Moraes of Brazil’s Supreme Court, known for his strict oversight of tech companies, reiterated that social media firms must adhere to Brazilian laws to continue operating in the country. Last year, he temporarily suspended X (formerly Twitter) over non-compliance with local regulations.

Meta has so far declined to comment on the matter in Brazil, fueling concerns over its commitment to tackling misinformation globally. The outcome of Brazil’s inquiry could have broader implications for how tech firms balance local laws with global policy changes.

Rounded enables businesses to build AI-powered voice assistants

French startup Rounded is developing an orchestration platform that allows companies to create their own AI voice agents. Initially focused on web3, the company shifted its attention to AI-powered voice interactions in mid-2023. Its first product, Donna, was designed for anesthetists, helping private hospitals handle large volumes of routine patient calls. The AI agent has already managed hundreds of thousands of conversations, improving in speed and accuracy over time.

After refining its technology, Rounded expanded its focus to offer a platform where businesses can build their own AI voice agents. Users can integrate various AI models, such as speech-to-text and text-to-speech engines, selecting components from providers like Azure, GPT-4o mini, and ElevenLabs. The platform also helps define prompts and parameters to optimise each agent’s performance for specific use cases.

The startup has secured €600,000 in funding from UC Berkeley’s SkyDeck accelerator and business investors. With growing interest in AI-powered customer interactions, Rounded is poised to attract further investment as it expands its product offering.

Google introduces AI-powered ‘Daily Listen’ podcast feature

Google is testing a new feature called “Daily Listen,” which generates personalised AI-powered podcasts based on users’ Discover feeds. The feature, currently rolling out to US users in the Search Labs experiment, provides a five-minute audio summary of topics tailored to individual interests. Each podcast includes links to related stories, allowing listeners to explore subjects in greater depth.

The experience is integrated with Google’s Discover and Search tools, using followed topics to refine content recommendations. Daily Listen functions similarly to NotebookLM’s Audio Overviews, which create AI-generated audio summaries based on shared documents. Users who have access to the feature will see a “Daily Listen” card on their Google app’s home screen, displaying a play button and episode length.

Once launched, the podcast plays alongside a rolling transcript, offering a seamless blend of text and audio. Google aims to enhance how users consume news and stay informed, making the experience more interactive and personalised. The feature reflects the company’s ongoing push into AI-driven content delivery.

Uttar Pradesh partners with Google Cloud to revolutionise agriculture with open digital network

The Government of Uttar Pradesh, a state in northern India, and Google Cloud have partnered to launch a pioneering open network for agriculture, powered by Google’s Gemini and enabled by the Beckn Protocol. The initiative will provide millions of farmers with seamless access to essential services, including advisory, credit, mechanisation, and market linkages, all through a single platform.

The network is designed to be open and decentralised, allowing any service provider—from input suppliers to financial institutions and government agencies—to connect with farmers. Google’s Gemini framework facilitates easy access by enabling voice command interactions in multiple languages, including Hindi, Bengali, Telugu, Kannada, Gujarati, and Punjabi, with more languages to be added in the future.

The Beckn Protocol ensures interoperability and decentralisation, making the network an innovative alternative to traditional, closed-market systems. It encourages the development of customised solutions, fostering competition and innovation within the agricultural ecosystem.

The open network is part of Uttar Pradesh’s broader vision to digitise agriculture, double farmers’ incomes, and bridge the digital divide using AI technologies. The initiative also serves as a model for other regions, demonstrating how digital public infrastructure (DPI) can transform agriculture on a large scale.

The Uttar Pradesh government aims to empower farmers with the tools and resources they need to thrive in the 21st century. By leveraging the Beckn Protocol and Google Cloud’s DPI-in-a-box solution, the network enables the decentralised exchange of value, making agri-commerce and advisory services more accessible. Through this collaboration, Uttar Pradesh seeks to accelerate progress in agriculture while fostering innovation and enhancing its farming community’s economic and social impact.