Google has unveiled VaultGemma, a new large language model built to offer cutting-edge privacy through differential privacy. The 1-billion-parameter model is based on Google’s Gemma architecture and is described as the most powerful differentially private LLM to date.
Differential privacy adds mathematical noise to data, preventing the identification of individuals while still producing accurate overall results. The method has long been used in regulated industries, but has been challenging to apply to large language models without compromising performance.
VaultGemma is designed to eliminate that trade-off. Google states that the model can be trained and deployed with differential privacy enabled, while maintaining comparable stability and efficiency to non-private LLMs.
This breakthrough could have significant implications for developers building privacy-sensitive AI systems, ranging from healthcare and finance to government services. It demonstrates that sensitive data can be protected without sacrificing speed or accuracy.
Google’s research teams say the model will be released with open-source tools to help others adopt privacy-preserving techniques. The move comes amid rising regulatory and public scrutiny over how AI systems handle personal data.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
South Australian public schools will soon gain access to EdChat, a ChatGPT-style app developed by Microsoft in partnership with the state government. Education Minister Blair Boyer said the tool will roll out next term across public high schools following a successful trial.
Safeguards have been built into EdChat to protect student data and alert moderators if students type concerning prompts, such as those related to self-harm or other sensitive topics. Boyer said student mental health was a priority during the design phase.
Teachers report that students use EdChat to clarify instructions, get maths solutions explained, and quiz themselves on exam topics. Adelaide Botanic High School principal Sarah Chambers described it as an ‘education equaliser’ that provides students with access to support throughout the day.
While many educators in Australia welcome the rollout, experts warn against overreliance on AI tools. Toby Walsh of UNSW said students must still learn how to write essays and think critically, while others noted that AI could actually encourage deeper questioning and analysis.
RMIT computing expert Michael Cowling said generative AI can strengthen critical thinking when used for brainstorming and refining ideas. He emphasised that students must learn to critically evaluate AI output and utilise the technology as a tool, rather than a substitute for learning.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
US electricity prices are rising as the energy demands of data centres surge, driven by the rapid growth of AI technologies. The average retail price per kilowatt-hour increased by 6.5% between May 2024 and May 2025, with some states experiencing significantly sharper increases.
Maine saw the sharpest rise in electricity prices at 36.3%, with Connecticut and Utah following closely behind. Utilities are passing on infrastructure costs, including new transmission lines, to consumers. In Northern Virginia, residents could face monthly bill increases of up to $37 by 2040.
Analysts warn that the shift to generative AI will lead to a 160% surge in energy use at data centres by 2030. Water use is also rising sharply, as Google reported its facilities consumed around 6 billion gallons in 2024 alone, amid intensifying global AI competition.
Tech giants are turning to alternative energy to keep pace. Google has announced plans to power data centres with small nuclear reactors through a partnership with Kairos Power, while Microsoft and Amazon are ramping up nuclear investments to secure long-term supply.
President Donald Trump has pledged more than $92 billion in AI and energy infrastructure investments, underlining Washington’s push to ensure the US remains competitive in the AI race despite mounting strain on the grid and water resources.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
AI can help build individual learning paths for Ukraine’s 3.5 million students, but its use must remain ethical, First Deputy Minister of Education and Science Yevhen Kudriavets has said.
Speaking to UNN, Kudriavets stressed that AI can analyse large volumes of information and help students acquire the knowledge they need more efficiently. He said AI could construct individual learning trajectories faster than teachers working manually.
He warned, however, that AI should not replace the educational process and that safeguards must be found to prevent misuse.
Kudriavets also said students in Ukraine should understand the reasons behind using AI, adding that it should be used to achieve knowledge rather than to obtain grades.
The deputy minister emphasised that technology itself is neutral, and how people choose to apply it determines whether it benefits education.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
Real-time translation is becoming a standard feature across consumer tech, with Samsung, Google, and Apple all introducing new tools. Apple’s recently announced Live Translation on AirPods demonstrates the utility of such features, particularly for travellers.
YouTube has joined the trend, expanding its multi-language audio feature to millions of creators worldwide. The tool enables creators to add dubbed audio tracks in multiple languages, powered by Google’s Gemini AI, replicating tone and emotion.
The feature was first tested with creators like MrBeast, Mark Rober, and Jamie Oliver. YouTube reports that Jamie Oliver’s channel saw its views triple, while over 25% of the watch time came from non-primary languages.
Mark Rober’s channel now supports more than 30 languages per video, helping creators reach audiences far beyond their native markets. YouTube states that this expansion should make content more accessible to global viewers and increase overall engagement.
Subtitles will still be vital for people with hearing difficulties, but AI-powered dubbing could reduce reliance on them for language translation. For creators, it marks a significant step towards making content truly global.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
South Korea and NATO have pledged closer cooperation on cybersecurity following high-level talks in Seoul this week, according to Yonhap News Agency.
The discussions, led by Ambassador for International Cyber Affairs Lee Tae Woo and NATO Assistant Secretary General Jean-Charles Ellermann-Kingombe, focused on countering cyber threats and assessing risks in the Indo-Pacific and Euro-Atlantic regions.
Launched in 2023, the high-level cyber dialogue aims to deepen collaboration between South Korea and NATO in the cybersecurity domain.
The meeting followed talks between Defence Minister Ahn Gyu-back and NATO Military Committee chair Giuseppe Cavo Dragone during the Seoul Defence Dialogue earlier this week.
Dragone said cooperation would expand across defence exchanges, information sharing, cyberspace, space, and AI as ties between Seoul and NATO strengthen.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
Alibaba has open-sourced its latest AI model, Qwen3-Next, claiming it is ten times more powerful and cheaper to train than its predecessor.
Developed by Alibaba Cloud, the 80-billion-parameter model reportedly performs on par with the company’s flagship Qwen3-235B-A22B while remaining optimised for deployment on consumer-grade hardware.
Qwen3-Next introduces innovations such as hybrid attention for long text processing, high-sparsity mixture-of-experts architecture, and multi-token prediction strategies. These upgrades boost both efficiency and model stability during training.
Alibaba also released Qwen3-Next-80B-A3B-Thinking, a reasoning-focused model that outperformed its own Qwen3-32B-Thinking and Google’s Gemini-2.5-Flash-Thinking in benchmark tests.
The release strengthens Alibaba’s position as a major player in open-source AI, following last week’s preview of its 1-trillion-parameter Qwen-3-Max model, which ranked sixth on UC Berkeley’s LMArena leaderboard.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
The US Federal Trade Commission has launched an inquiry into AI chatbots that act as digital companions, raising concerns about their impact on children and teenagers.
Seven firms, including Alphabet, Meta, OpenAI and Snap, have been asked to provide information about how they address risks linked to ΑΙ chatbots designed to mimic human relationships.
Chairman Andrew Ferguson said protecting children online was a top priority, stressing the need to balance safety with maintaining US leadership in AI. Regulators fear minors may be particularly vulnerable to forming emotional bonds with AI chatbots that simulate friendship and empathy.
An inquiry that will investigate how companies develop AI chatbot personalities, monetise user interactions and enforce age restrictions. It will also assess how personal information from conversations is handled and whether privacy laws are being respected.
Other companies receiving orders include Character.AI and Elon Musk’s xAI.
The probe follows growing public concern over the psychological effects of generative AI on young people.
Last month, the parents of a 16-year-old who died by suicide sued OpenAI, alleging ChatGPT provided harmful instructions. The company later pledged corrective measures, admitting its chatbot does not always recommend mental health support during prolonged conversations.
Would you like to learn more aboutAI, tech and digital diplomacy? If so, ask our Diplo chatbot!
The US House of Representatives has passed an $848 billion defence policy bill with new provisions for cybersecurity and AI. Lawmakers voted 231 to 196 to approve the chamber’s version of the National Defence Authorisation Act (NDAA).
The bill mandates that the National Security Agency brief Congress on plans for its Cybersecurity Coordination Centre and requires annual reports from combatant commands on the levels of support provided by US Cyber Command.
It also calls for a software bill of materials for AI-enabled technology that the Department of Defence uses. The Pentagon will be authorised to create up to 12 generative AI projects to improve cybersecurity and intelligence operations.
An adopted amendment allows the NSA to share threat intelligence with the private sector to protect US telecommunications networks. Another requirement is that the Pentagon study the National Guard’s role in cyber response at the federal and state levels.
Proposals to renew the Cybersecurity Information Sharing Act and the State and Local Cybersecurity Grant Program were excluded from the final text. The Senate is expected to approve its version of the NDAA next week.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
The contact details of rail passengers have been stolen in a cyberattack affecting London North Eastern Railway (LNER). The company stated that it had been notified of unauthorised access to files managed by a third-party supplier and advised customers to be vigilant against phishing attempts.
LNER stressed that no bank details, card numbers, or passwords had been compromised. The York-based operator stated that it was collaborating with cybersecurity experts and the supplier to investigate the breach and ensure necessary safeguards.
The company did not confirm the number of passengers affected. The incident comes as LNER reported revenues exceeding £1 billion, yet it continues to rely on government support since its nationalisation in 2018.
Passenger complaints rose 12.2 percent in 2025, reaching 24,015, and competition from private operators is driving losses—online ticket platforms such as Trainline direct passengers to cheaper rivals, costing LNER significant revenue.
The breach follows other attacks on UK transport services, including a 2024 incident in which the bank details of 5,000 Transport for London customers were exposed, resulting in weeks of disrupted online services.
Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!
