Artificial intelligence

Anthropic boosts cloud capacity with Google’s AI hardware

Anthropic has struck a multibillion-dollar deal with Google to expand its use of cloud computing and specialised AI chips. The agreement includes the purchase of up to one million Tensor Processing Units, Google’s custom hardware built to train and run large AI models.

The partnership will provide Anthropic with more than a gigawatt of additional computing power by late 2026. Executives said the move will support soaring demand for its Claude model family, which already serves over 300,000 business clients.

Anthropic, founded by former OpenAI employees, has quickly become a major player in generative AI. Backed by Amazon and valued at $183 billion, the company recently launched Claude Sonnet 4.5, praised for its coding and reasoning abilities.

Google continues to invest heavily in AI hardware to compete with Nvidia’s GPUs and rival US tech giants. Analysts said Anthropic’s expansion signals intensifying demand for computing power as companies race to lead the global AI revolution.

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

Two founders turn note-taking into an AI success

Two 20-year-old drop-outs, Rudy Arora and Sarthak Dhawan, are behind Turbo AI, an AI-powered notetaker that has grown to around 5 million users and reached a multi-million-dollar annual recurring revenue (ARR) in a short timeframe.

Their app addresses a clear pain point, which is that meetings, lectures, and long videos produce information overload. Turbo AI uses generative AI to convert audio, typed notes or uploads into structured summaries, highlight key points and help users organise insights. The founders describe it as a ‘productivity assistant’ more than a general-purpose chat agent.

The business model appears lean, meaning that freemium user acquisition is scaling quickly, then converting power users into paid subscriptions. The insights are that a well-targeted niche tool can win strong uptake even in a crowded productivity-AI market.

Arora and Dhawan say they kept the feature set focused and user experience simple, enabling rapid word-of-mouth growth.

The growth raises interesting implications for enterprise and consumer AI alike. While large language models dominate headlines, tools like Turbo AI show the value of vertical-specific applications addressing tangible workflows (e.g., note-taking, summarisation). It also underscores how younger founders are building AI tools outside the major tech hubs and scaling globally.

At this stage, challenges remain: user retention, differentiation in a field where major players (Microsoft, Google, OpenAI) are adding similar capabilities, and privacy/data governance (especially with audio and meeting content). However, the early results suggest that targeted AI productivity tools can achieve a meaningful scale quickly.

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

Tech giants push AI agents into web browsing

Tech companies are intensifying competition to reshape how people search online through AI-powered browsers. OpenAI’s new Atlas browser, built around ChatGPT, can generate answers and complete web-based tasks such as making shopping lists or reservations.

Atlas joins rivals like Microsoft’s Copilot-enabled Edge, Perplexity’s Comet, and newer platforms Dia and Neon. Developers are moving beyond traditional assistants, creating ‘agentic’ AI capable of acting autonomously while keeping user experience familiar.

Google remains dominant, with Chrome holding over 70 percent of the browser market and integrating limited AI features. Analysts say OpenAI could challenge that control by combining ChatGPT insights with browser behaviour to personalise search and advertising.

Experts note the battle extends beyond browsers as wearables and voice interfaces evolve. Controlling how users interact with AI today, they argue, could determine which company shapes digital habits in the coming decade.

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

AI tool Mirror keeps track of medical information

A new app called Mirror, developed by Oxford-based company Aide Health, aims to help patients remember and summarise information from medical appointments using AI. The platform records consultations and produces summaries that patients can refer back to or share with family and carers.

Creator Ian Wharton said the idea came from helping his father, who has early-stage Alzheimer’s, to recall essential details from doctors’ visits. The app listens passively during appointments and produces a clear summary of what was discussed, making it easier for patients to retain key information.

Early users have praised the platform for making consultations easier to manage. One described being able to share concise summaries with friends and colleagues, saving the effort of repeating complex medical details. The creator added that patient data is private and not shared with third parties.

The current version works during in-person consultations, but future updates will allow the app to actively prompt patients with reminders or questions, advocating for their healthcare needs.

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

MoonshotAI released KIMI-K2 and OK Computer

KIMI-K2 is a large language model (LLM) developed by Beijing-based Moonshot AI, offering strong performance in writing and coding across diverse applications. Open-source and versatile, it delivers high-quality outputs across multiple domains, from text generation to programming.

Alongside KIMI-K2, the developers introduced OK Computer, an agent that extends the model’s abilities. Using this agent, users can build websites, conduct research, generate images, and create presentations or graphics from a single prompt, making complex workflows simpler and more accessible.

These tools reflect a growing trend in AI, which is combining multiple capabilities into one accessible system. By offering open-source solutions, KIMI-K2 and OK Computer empower users to tackle creative, technical, and research tasks with minimal effort.

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

MLK estate pushback prompts new Sora 2 guardrails at OpenAI

OpenAI paused the ability to re-create Martin Luther King Jr. in Sora 2 after Bernice King objected to user videos. Company leaders issued a joint statement with the King estate. New guardrails will govern depictions of historical figures on the app.

OpenAI said families and authorised estates should control how likenesses appear. Representatives can request removal or opt-outs. Free speech was acknowledged, but respectful use and consent were emphasised.

Policy scope remains unsettled, including who counts as a public figure. Case-by-case requests may dominate early enforcement. Transparency commitments arrived without full definitions or timelines.

Industry pressure intensified as major talent agencies opted out of clients. CAA and UTA cited exploitation and legal exposure. Some creators welcomed the tool, showing a split among public figures.

User appetite for realistic cameos continues to test boundaries. Rights of publicity and postmortem controls vary by state. OpenAI promised stronger safeguards while Sora 2 evolves.

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

Church of Greece launches AI tool LOGOS for believers

LOGOS, a digital tool developed by the Metropolis of Nea Ionia, Filadelfia, Iraklio and Halkidona alongside the University of the Aegean, has marked the Church of Greece’s entry into the age of AI.

The tool gathers information on questions of Christian faith and provides clear, practical answers instead of replacing human guidance.

Metropolitan Gabriel, who initiated the project, emphasised that LOGOS does not substitute priests but acts as a guide, bringing believers closer to the Church. He said the Church must engage the digital world, insisting that technology should serve humanity instead of the other way around.

An AI tool that also supports younger users, allowing them to safely access accurate information on Orthodox teachings and counter misleading or harmful content found online. While it cannot receive confessions, it offers prayers and guidance to prepare believers spiritually.

The Church views LOGOS as part of a broader strategy to embrace digital tools responsibly, ensuring that faith remains accessible and meaningful in the modern technological landscape.

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

OpenAI rolls out pet-centric AI video features and social tools in Sora

OpenAI has announced significant enhancements to its text-to-video app Sora. The update introduces new features including pet and object ‘cameos’ in AI-generated videos, expanded video editing tools, social sharing elements and a forthcoming Android version of the app.

Using the new pet cameo feature, users will be able to upload photos of their pets or objects and then incorporate them into animated video scenes generated by Sora. The objective is to deepen personalisation and creative expression by letting users centre their own non-human characters.

Sora is also gaining editing capabilities that simplify the creation process. Users can remix existing videos, apply stylistic changes, and integrate social-type features like feeds where others’ creations can be viewed and shared. The Android app is noted as ‘coming soon’ which expands Sora’s accessibility beyond the iOS/web initial release.

The move reflects OpenAI’s strategy to transition Sora from an experimental novelty into a more fully featured social video product. By enabling user-owned content (pets/objects), expanding sharing functionality and broadening platform reach, Sora is positioned to compete in the generative video and social media landscape.

At the same time, the update raises questions around content use, copyright (especially when user-owned pets or objects are included), deepfake risks, and moderation. Given Sora’s prior scrutiny over synthetic media, the expansion into more personalised video may prompt further regulatory or ethical review.

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

Meta cuts 600 AI roles even as it expands superintelligence lab

Meta Platforms confirmed today it will cut approximately 600 jobs from its AI division, affecting teams including the Fundamental AI Research (FAIR) unit and product and infrastructure units. The move comes even as the company continues hiring for its elite superintelligence unit, the TBD Lab, which remains unaffected by the cuts.

According to an internal memo from Chief AI Officer Alexandr Wang, the layoff aim is to make remaining teams more load-bearing and impactful. ‘By reducing the size of our team, fewer conversations will be required to make a decision, and each person will be more load-bearing and have more scope and impact,’ Wang wrote.

Meta says employees affected will be encouraged to apply for other roles within the company; many are expected to be reassigned. The company’s earlier hiring spree in AI included poaching top talent from competitors and investing heavily in infrastructure. Analysts say the current cuts reflect a strategic pivot rather than a retreat, from broad AI research to more focused, high-impact model development.

This shift comes as Meta competes with organisations like OpenAI and Google in the race to build advanced large-language models and scaled AI systems. By trimming staff in legacy research and infrastructure units while bolstering resources for its superintelligence arm, Meta appears to be doubling-down on frontier AI even as it seeks to streamline operations.

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

Diella 2.0 set to deliver 83 new AI assistants to aid Albania’s MPs

Albania’s AI minister Diella will ‘give birth’ to 83 virtual assistants for ruling-party MPs, Prime Minister Edi Rama said, framing a quirky rollout of parliamentary copilots that record debates and propose responses.

Diella began in January as a public-service chatbot on e-Albania, then ‘Diella 2.0’ added voice and an avatar in traditional dress. Built with Microsoft by the National Agency for Information Society, it now oversees specific state tech contracts.

The legality is murky: the constitution of Albania requires ministers to be natural persons. A presidential decree left Rama’s responsibility to establish the role and set up likely court tests from opposition lawmakers.

Rama says the ‘children’ will brief MPs, summarise absences, and suggest counterarguments through 2026, experimenting with automating the day-to-day legislative grind without replacing elected officials.

Reactions range from table-thumping scepticism to cautious curiosity, as other governments debate AI personhood and limits; Diella could become a template, or a cautionary tale for ‘ministerial’ bots.

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