As organizations move AI from experimentation to production, governance requirements...

Discussion | Link

PLUS: Conduct better stock research with Perplexity Finance

Jim Franke pulls away the cover page of a presentation on the wraparound desk in his office, revealing an illustration of an odd-­looking aircraft with massive wings stretching out from a stubby fuselage. The uncrewed plane is soaring thousands of meters higher than commercial jets fly—so high you can see the curvature of the Earth.…

Most of Kenya’s power grid runs on renewables. But with 25% of communities lacking centralized electricity, the nation is looking to off-grid solar to hit its goal of delivering universal electricity access by 2030 without driving up emissions. The ever-­improving economics of solar technology have helped. A couple of years ago, a panel cost about…

Financial Times: World models startup Odyssey raised $310M from Amazon and others at a $1.45B valuation and will use AWS as its preferred cloud partner, deploying Trainium chips  —  Company joins investment arms of Nvidia and AMD in $310mn funding round for Odyssey ML  —  Amazon is partnering with a start …

There’s never been a more dynamic, exciting time to be building your own AI models...

The UK will ban adolescents under 16 years old from user-to-user social media platforms, despite age verification issues and privacy concerns.

Othman Alghanmi / Unsplash Last Friday, US-based artificial intelligence (AI) company Anthropic received an “export control” directive from its government. The company was told it must block access to two of its most capable models, Fable and Mythos, for all foreign nationals. Within hours, Anthropic shut down access to the models for users everywhere in the world, including researchers, clinicians and analysts in Australia. This happened with no warning and no backup plan. Why did this happen? The directive’s “foreign national” criterion is a citizenship concept. However, Anthropic and other cloud-based AI providers only know the location of their users, not their citizenship. Consumer AI services have no effective mechanism to verify citizenship. Even their location filters can be dodged with tools such as virtual private networks (VPNs). There is no way any control of these services based on user nationality can be enforced. A system that blocks a foreign national in the United States but allows access to a US national in Australia, for example, is not currently available. The US government directive was a geopolitical signal – an indication rather than an enforceable control. The only way Anthropic could comply was to shut down access for people everywhere. A question of sovereignty Australian universities, government agencies, health systems and industry have integrated US-hosted frontier AI deeply into their operations. Advanced analytics platforms, AI-assisted research tools, and productivity infrastructure built on top of models such as Claude or GPT-5 operate with an implicit assumption: that access will persist. The same is true of ubiquitous systems such as Microsoft Teams or Google Workspace, but as we are seeing the assumption may not hold for AI. The Anthropic shutdown was a reminder that access and control sat entirely within US jurisdiction, regardless of where any Australian user happened to be. It represents a failure of data and technology sovereignty: our ability to operate without permission from other nations. Australia’s exposure The Anthropic shutdown is the first legally enforceable export control targeting a software-level AI system. It will not be the last. Any frontier AI model hosted in the US, by Anthropic, OpenAI, Google DeepMind or Meta, is subject to US export control law. The precedent set this week extends to all of them. There is some debate about whether an export control only applies to physical exports and not remote access to models housed in the US, and some experts have suggested the order may be challenged on those grounds. If that were to happen, we would expect the US government to change the regulations. The US government has previously used technology export controls as a geopolitical instrument, particularly in the case of chips and semiconductors. Even US allies such as Australia are not exempt from these controls. The Anthropic order should have come as no surprise. Voices in Australia’s own research community including Jon Whittle, former head of CSIRO Data61, had been publicly warning about exactly this scenario for more than a year. To prepare for the future, Australia needs a strategy for its own sovereign AI. This can’t be a distant aspiration: it needs to be an operational plan with named owners, timelines and budget. How does Australia achieve AI sovereignty? A true sovereign AI strategy requires four things. First is data sovereignty: data that is physically stored in Australia and subject to Australian law. Second is compute sovereignty: in-country data centres under Australian control. Third, AI model sovereignty: AI capability that is not dependent on a foreign provider. And finally, policy sovereignty: the ability to set its own rules, rather than inheriting another country’s export controls or safety regulations by default. Australia already has some examples of sovereign data and compute, such as Macquarie Government, Vault Cloud and AUCyber. Building this capacity will require huge, sovereign data centres on Australian soil. However, that build-out will only garner public support if it comes with binding commitments on renewable energy sourcing, water-efficient cooling, and transparent community consultation in regions where data centres are proposed (such as Bundey in South Australia). For Australian residents, this sovereignty means not depending on AI tools and systems that another nation can switch off, restrict or alter at will. Developing sovereign AI models will be a challenge. However, it doesn’t necessarily mean building new systems from scratch. Switzerland’s Apertus model, released in September 2025, is an excellent example of what open-source approaches can produce. It also shows how large language models can be made to comply with regulations such as the European Union’s Artificial Intelligence Act. Beyond this, sovereign capability will also require training future experts. We can expect increased demand for people who can build the technology, deploy and manage it, govern it, and integrate it into existing systems. The regulatory dimension The EU is developing AI policy frameworks from which Australia can learn. It is establishing reciprocal access rights with technology partners and doing institutional contingency planning – and putting money into the process. Several European countries are moving rapidly. France has just announced a further €655 million (A$1 billion) going into national AI capabilities, explicitly referencing last week’s events. This builds on roughly €2.5 billion of investment since 2018. Sovereign AI does not mean reinventing the frontier model from scratch. It can mean public, transparent, nationally governed infrastructure built on open foundations. At a minimum, AI sovereignty requires making sure we are not too dependent on single AI suppliers or even suppliers based in a single jurisdiction. The goal should be that no single government or jurisdiction can unilaterally cut off Australian residents and institutions. Dimitri Perrin has received funding from the Australian Research Council (ARC), the National Health and Medical Research Council (NHMRC), the Australian Research Data Commons (ARDC), the Australian-French Association for Innovation and Research (AFRAN), and the Advance Queensland programme. Bernadette Hyland-Wood and Michael Guihot do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have disclosed no relevant affiliations beyond their academic appointment.

Wall Street Journal: A profile of Anthropic researcher Nicholas Carlini, who warned about Mythos in March and is now part of an Anthropic team briefing the White House on safeguards  —  Nicholas Carlini recently rang the alarm about the dangers of AI—and now he's part of a team arguing for the latest models to be released

Nikkei Asia: Sources: Samsung is seeing a rise in advanced chip production requests from BYD, Google, AMD, and others as TSMC's capacity remains constrained by AI demand  —  TAIPEI/HONG KONG/SEOUL — BYD, Google, AMD, Tesla and others are looking increasingly to Samsung Electronics for contract chipmaking services …

Nikhil Pahwa / @nixxin: India's Telegram ban until June 22 over exam paper leak concerns is a terrible look for the country, sets a precedent, and could happen again if unchallenged  —  A few additional thoughts on the Telegram ban in India: 1. Is this a permanent ban? It's unusual for the Indian government to issue a press release about the telegram ban: the precedence that I remember is when Chinese apps were banned in India in June 2020 including tiktok. A

The Admiral Grigorovich, pictured here in Sudan, has been sailing in the English Channel as Russia's shadow fleet tankers come under threat from seizure by the UK. IBRAHIM ISHAQ/AFP via Getty Images Russia said one of its warships fired warning shots at a British civilian yacht on Sunday. The frigate Admiral Grigorovich was in the English Channel when it opened fire with small arms. A retired British couple said they were sailing on the yacht when they encountered the frigate. A Russian frigate opened fire in the English Channel on Tuesday, firing warning shots with small arms near a UK-registered civilian yacht, London and Moscow said. The warship Admiral Grigorovich fired several shots — single rounds, not automatic fire — near the Bright Future, a sailing yacht, roughly 20 nautical miles south of the Isle of Wight outside UK territorial waters. The UK assesses that the Grigorovich was signaling to other vessels that it was drifting instead of maneuvering under power, possibly leaving the warship feeling vulnerable. It sounded warnings before opening fire. "Following attempts to contact a British vessel in the channel, the Grigorovich fired warning shots," a UK defense ministry spokesperson told Business Insider. "These were not aimed at the vessel and were an attempt to prevent a possible collision." Russia's defense ministry said the frigate had attempted to contact the Bright Future with radio, signal flares, and sound signals, but opened fire after receiving no response and seeing the yacht "following a dangerous course." "After closing the distance to 150 meters, the frigate's commander decided to carry out the preemptive fire at the vessel's course with small arms," Moscow said. A retired British couple on board the Bright Future told the BBC that the two vessels were not on a collision course and that the yacht had adjusted its path after the Admiral Grigorovich issued five horn blasts. The incident follows the UK's Royal Marines' separate seizure of the MV Smyrtos, a tanker believed to be part of Russia's shadow fleet, off the southern coast of England on Sunday. Military helicopters boarded the MV Smyrtos off the coast of Portland. Finnbarr Webster/Getty Images The boarding operation was a first for the UK, involving two British warships, reconnaissance aircraft, and helicopters. While both events occurred in the English Channel, the UK defense ministry said that the seizure and Tuesday's warning shots from the Admiral Grigorovich were isolated incidents. "HMS Mersey has been monitoring the Russian vessel, and support has been provided to the crew of the yacht," the defense ministry spokesperson said. Still, the Russian navy has been repeatedly reported to be escorting shadow fleet tankers in convoys. The Admiral Grigorovich, part of the Black Sea Fleet, was spotted convoying two tankers in the English Channel in April, just after UK Prime Minister Keir Starmer had given British forces the authority to seize shadow fleet vessels. Retired Royal Navy Commodore Steve Prest, now an associate fellow at the UK's Royal United Services Institute, said it is possible that the Grigorovich's commanding officer decided to open fire after they got too nervous about an unresponsive yacht near the warship. However, Prest said in comments shared with Business Insider, given the context of the shadow fleet and the Royal Marines' seizure of the Smyrtos, "I think this is the Russians baring their teeth," adding that Moscow "very rarely will do something like this in an uncalculated, haphazard way." Prest said Russia may be trying to signal to other ships preparing to go through the English Channel: "Hey, look, we are here, we are serious, and we are prepared to stand our ground, so let's not have any miscalculation." The 409-foot-long guided-missile frigate is the lead ship of its class and was commissioned in 2016. The ship's main armament consists of eight vertical launch cells for land-attack cruise missiles and a 100mm naval gun. June 16, 2026: This story was updated to reflect comment from the UK Defense Ministry. Read the original article on Business Insider

Discussion | Link

The Information: Sources: a new AI lab founded by Junyang Lin, former lead of Alibaba's Qwen models, has raised several hundred million dollars at a ~$2B post-money valuation  —  Tencent Holdings has backed a new AI lab founded by Junyang Lin, former lead researcher of Alibaba Group's Qwen models …

Financial Times: Dealogic: investors have poured $58B across 42 data center deals YTD; Oxford Economics says ~850 data centers worth some $7T are under construction globally  —  Sector remains hot despite some investors' wariness about tenants such as ByteDance  —  Data centre deals are weathering a storm …

Getty Images; Tyler Le/BI Vibe coding companies are picking up big money and even some controversy. Lovable and Replit are seeing valuations soar even as competition grows. Cursor just sold to SpaceX for $60 billion. The space has seen lots of deal activity and interest from Big Tech. The tech world is both in awe of and fearful of vibe coding. On one hand, tech giants are all in on these AI-assisted coding tools. They're touting efficiency gains, listing it as a need-to-have in job descriptions, buying their employees subscriptions, and even investing in vibe-coding startups themselves. In the latest news from the vibe-coding bonanza: SpaceX said on Tuesday that it was officially acquiring AI coding startup Cursor for $60 billion. The deal strengthens SpaceX's position in the AI coding race by helping it compete with top labs building advanced coding tools. That deal is just the latest in a broader wave of acquisitions and partnerships sweeping through the vibe-coding space. In July, AI startup Cognition snatched up Windsurf after OpenAI's $3 billion deal to acquire the vibe coding tool maker fell through. Just one month earlier, web design platform Wix bought Base44, a six-month-old startup bootstrapped by a solo founder, for $80 million. These entrants are competing with far bigger and better-funded players, including OpenAI, Anthropic, and Microsoft, that make their own AI-powered coding tools. Yet, the rise of these tools is also rattling the broader market: Some tech giants saw their shares take a hit as investors dumped legacy software stocks over concerns that AI and vibe coding will allow companies to build their own software rather than buy. Both narratives are driving the valuations of vibe-coding startups such as Lovable and Replit, now well into the billions. "Our mission has always been that every human with an idea and an internet connection should be able to build any app they want," Amjad Masad, the CEO of Replit, said in a release in March, announcing his company's $9 billion valuation. Business Insider compiled a list of the startups riding the vibes, detailing their latest valuations, fundraises, and what they're best known for. Have a tip? Contact Shubhangi Goel via email at sgoel@businessinsider.com or Signal at shuby.85. Use a personal email address and a nonwork device; here's our guide to sharing information securely. Lovable Lovable CEO Anton Osika. Bruno de Carvalho/SOPA Images/LightRocket via Getty Images Based in Stockholm and launched in 2024, Lovable is among the biggest players in the vibe coding world and one of the fastest-growing startups. In March, Business Insider reported that the Swedish startup's annual recurring revenue had surged by more than 30%, from $300 million to $400 million in a single month. ARR, a key metric used to gauge startup performance, refers to the predictable revenue a company expects to generate over a year. Lovable, founded by Anton Osika and Fabian Hedin, was valued at $6.6 billion in a December funding round led by CapitalG and Menlo Ventures. Lovable's chief revenue officer, Ryan Meadows, told Business Insider that the company plans to more than double its head count by the end of the year, from 146 to 350 employees. He added that Lovable, which specializes in making coding user-friendly, now sees 200,000 new vibe-coding projects created each day. Replit Replit CEO Amjad Massad. Stephen McCarthy/Sportsfile for Web Summit Qatar via Getty Images Replit, founded in 2016, touts itself as an all-in-one platform that not only generates code but also builds, hosts, and deploys applications in one place. Over the past few years, Replit has pivoted from a collaborative coding environment to the Replit Agent that can turn plain-English descriptions into working applications, lowering the barrier to entry for beginner coders. In March, the startup announced it had raised a $400 million Series D round at a $9 billion valuation, led by its previous investor, Georgian Partners. Other investors include Coatue, Andreessen Horowitz, Craft Ventures, Accenture Ventures, and angels Shaquille O'Neal and Jared Leto. On May 28, Visa announced it had invested an undisclosed amount in Replit as part of a partnership. Emergent Emergent Emergent, founded out of Y Combinator's startup class of 2024 by twin brothers Mukund Jha and Madhav Jha, is one of the newest but fastest-growing vibe coding platforms. Similar to Replit, Emergent says it allows users to "build full-stack, production-ready applications using just natural language prompts." The startup said in February that it had 6 million users and had reached $100 million in ARR in eight months. Its latest funding round, raised in January, raised $70 million in Series B funding from Khosla Ventures and SoftBank Vision Fund 2, with participation from Prosus, Lightspeed, Together, and Y Combinator. The startup's valuation was not disclosed. Emergent's $23 million Series A round closed in September, signaling how eager investors are to get in on the growing pie. "A lot of the other platforms, they're great for prototyping, they're great for demos, but when it comes to really managing the entire lifecycle of software development, they fall short," CEO Mukund Jha told Business Insider. "That's a gap we are trying to fill in the market right now." Poolside AI Co-Founder and Chief Technology Officer of Poolside, Eiso Kant . JOEL SAGET/AFP via Getty Images San Francisco-based Poolside was cofounded in 2023 by former GitHub head of tech Jason Warner and software entrepreneur Eiso Kant. The company focuses on selling to enterprises and public sector organizations. It builds models that can write computer software and coding applications. In October, Bloomberg reported that the company was in discussions to raise $2 billion at a $12 billion valuation, with a potential $500 million to $1 investment from Nvidia. The company closed a $500 million Series B in 2024, led by Bain Capital, with participation from Nvidia, a Poolside representative told Business Insider. It is raising a Series C, and Nvidia has committed at least $500 million to anchor the round. Correction: March 13, 2026 — An earlier version of this story misstated Poolside's headquarters. The company is based in San Francisco, not Paris. StackBlitz's Bolt StackBlitz cofounders Albert Pai (left) and Eric Simons (right). Eric Simons StackBlitz, founded in 2017 and headquartered in San Francisco, credits its survival to Bolt, a vibe coding platform the company launched in 2024 when it was struggling with dwindling revenue. Bolt, which uses Anthropic's models to let users build what they want with plain English, generated about $1 million in ARR in the first week it came out, cofounder Eric Simons told Business Insider last year. The week after, it added another $1 million in ARR, and then another. "I had slept three hours a night for a week straight to get the release out with our team," Simons told Business Insider about Bolt's release. "After seeing it live, and people loving it — beyond anything I had ever created before — I cried, alone at my desk in my backyard shed office." In January 2025, Bloomberg reported that StackBlitz was in talks with investors to raise $83.5 million at a $700 million valuation. Cognition Scott Wu cofounded Cognition in 2023. IMF via Getty Images Cognition was founded in 2023 by former competitive programmers Scott Wu, Steven Hao, and Walden Yan. The company is best known for Devin, an autonomous AI software engineer that can plan, write, test, and deploy software. Devin is an AI teammate capable of managing the entire software development lifecycle, unlike tools that focus only on code generation, said Emily Cohen, who heads people and operations at the AI startup. The San Francisco-based company has attracted backing from investors including Founders Fund, 8VC, Lux Capital, Khosla Ventures, Elad Gil, and Pear VC. In May, the company raised more than $1 billion at a $26 billion post-money valuation, making it one of the most valuable AI coding startups globally. Cognition acquired AI coding platform Windsurf in 2025. It established offices in London and Singapore earlier this year. Kilo Kilo Code is a San Francisco-based AI coding startup founded in 2025 by Scott Breitenother, Emilie Schario, and Sid Sijbrandij, the cofounder and former CEO of GitLab. The company offers an open-source AI coding agent that helps developers write, edit, and manage software using natural-language prompts, while supporting access to hundreds of AI models. "What do you do if Anthropic has a down day? Do you say, 'Hey folks, everybody take Friday off.' I mean, you can't," Breitenother, Kilo's CEO, said in an interview with Business Insider. "That's where Kilo comes in. Sure, you can buy Anthropic, but also buy OpenAI and use them both through Kilo." In December, Kilo raised an $8 million seed round led by Cota Capital, with participation from Breakers, General Catalyst, Quiet Capital, Tokyo Black, and others. Elon Musk posted on X in June that Kilo's tools were "not bad for version 0.1" and "good value for money." SkipLabs AI infrastructure startup SkipLabs was founded in 2022 by Julien Verlaguet, the creator of Facebook's Hack programming language. In June, the company launched Skipper, a closed-loop coding agent for back-end services aimed at professional developers. In March 2025, SkipLabs raised an $8 million seed round led by Amplify Partners, with participation from Tapestry VC and angel investors including Yann LeCun, Spencer Kimball, and Olivier Pomel. Read the original article on Business Insider

The New York Knicks will be honored in Manhattan on Thursday with a ticker-tape parade. San Antonio Express-News/Hearst Newspapers/San Antonio Express-News via Getty Images The Knicks are set to celebrate their NBA championship with a ticker-tape parade in Manhattan on Thursday. The event has given rise to a mini gig economy, with people offering to hold spots and others seeking their services online. Multiple listings have cropped up on gig-work platform Airtasker, with asking prices ranging from $50 to $800. One of the hottest pieces of real estate in New York City this week isn't an apartment. It's a piece of Lower Manhattan sidewalk along the Knicks' championship parade route — and some people are charging hundreds of dollars to hold one. Ahead of Thursday's celebration, would-be line sitters are advertising their services online, offering to secure prime viewing spots as the procession winds from Bowling Green through the "Canyon of Heroes" to City Hall. On Airtasker, a gig-work platform where users can hire people for tasks and errands, multiple Knicks parade-related listings have cropped up, with asking prices ranging from $50 to $800. The listings come from both sides of the market: people offering to hold spots along the parade route, as well as fans looking to hire someone to wait in their place. A screengrab of the Airtasker website showing listings related to the Knicks' championship parade. Airtasker One listing offered $750 to someone who could secure three spots near Broadway and Barclay Street from midnight Wednesday until 8 a.m. Thursday, two hours before the parade is set to begin. The task had already been assigned by Tuesday night. The parade comes after the Knicks ended a 53-year championship drought on Saturday, defeating the San Antonio Spurs in Game 5 of the NBA Finals. "Through near misses, heartbreak and a hope that every year could be our year, this city never stopped believing in the Knicks. And this team fulfilled that hope with grit, resilience and heart — just like the five boroughs itself," New York City mayor Zohran Mamdani said in a statement on Saturday. Following the parade, the mayor will present the Knicks with the keys to the city during a championship ceremony at City Hall Plaza. The city is also giving away 600 tickets to the event through a public lottery that closes at 11 a.m. on June 17. Read the original article on Business Insider

Discussion | Link