A scale model in cork of First Church, Dunedin intended for the New Zealand and South Seas Exhibition. — Otago Witness, 30.6.1925 Although seventy-seven years is but a comparatively brief span in the history of the greater churches of the Homeland, there are but few churches in New Zealand which have rendered service to the community for over "three score years and ten."
It’s home. PHOTO: LINDA ROBERTSON Today's Letters to the Editor from readers cover topics including retirement villages or Ponzi schemes, phonics and learning to read, and Israelis who resist the killing in Gaza.
Over the next few years, agentic AI is expected to bring not only rapid technological breakthroughs, but a societal transformation, redefining how we live, work and interact with the world . And this shift is happening quickly. “By 2028, 33% of enterprise software applications will include agentic AI, up from less than 1% in 2024, enabling 15% of day-to-day work decisions to be made autonomously,” according to research firm Gartner. Unlike traditional AI, which typically follows preset rules or algorithms, agentic AI adapts to new situations, learns from experiences, and operates independently to pursue goals without human intervention. In short, agentic AI empowers systems to act autonomously, making decisions and executing tasks — even communicating directly with other AI agents — with little or no human involvement. One key driver is the growing sophistication of large language models (LLMs), which provide the “brains” for these agents. Agentic AI will enable machines to interact with the physical world with unprecedented intelligence, allowing them to perform complex tasks in dynamic environments, which could be especially useful for industries facing labor shortages or hazardous conditions.The rise of agentic AI also brings security and ethical concerns. Ensuring these autonomous systems operate safely, transparently and responsibly will require governance frameworks and testing. Follow this page for ongoing agentic AI coverage from Computerworld and Foundry’s other publications. Agentic AI news and insights Salesforce AI Research unveils new tools for AI agents August 27, 2025 : Salesforce announced a simulated enterprise environment, benchmark, and account data unification tool that are designed to help customers transform into agentic AI enterprises . Agentic AI promises a cybersecurity revolution — with asterisks August 18, 2025: The hottest topic at this year’s Black Hat conference was the meteoric emergence of AI tools for both cyber adversaries and defenders , particularly the use of agentic AI to strengthen cybersecurity programs. 4 thoughts on who should manage AI agents August 11, 2025: As AI agents proliferate, we need to turn our attention beyond AI agent builder platforms to AI orchestration and AI GRC platforms. It also raises questions about which groups within the enterprise should manage AI agents and how they should be treated. How bright are AI agents? Not very, recent reports suggest July 31, 2025 : Security researchers are adding more weight to a truth that infosec pros had already grasped: AI agents are not very bright , and are easily tricked into doing stupid or dangerous things Will AI agents eat the SaaS market? Experts are split July 31,2025 : As hype about AI agents reaches new heights, an emerging theory suggests that the groundbreaking AI tools will kill the SaaS business model. The claim isn’t particularly new, but is resurfacing, with people like Microsoft CEO Satya Nadella voicing this position. How agentic AI will change database management July 28, 2025: Generative AI has already had a profound impact on the world of database management. And now, thanks to AI’s knack for pattern-recognition, teams can use generative AI to analyze data sets, detect anomalies, and access invaluable insights with record speed and precision. As AI agents go mainstream, companies lean into confidential computing for data security July 21, 2025 : Companies need to stop ignoring data security as AI agents take over internal data movement in IT environments, analysts and IT execs warn. To address that issue, some tech players are embracing the concept of “confidential computing.” While it’s existed for years, it;s now finding new life with the rise of genAI. How agentic AI will transform mobile apps and field operations July 15, 2015 : Agentic AI will usher in new mobile AI experiences . Construction, manufacturing, healthcare, and other industries with significant field operations will benefit from mobile AI agents and the resulting operational agility. MCP is fueling agentic AI — and introducing new security risks July 10, 2025 : Model Context Protocol (MCP) has caught fire, with several thousand MCP servers now available from a wide range of vendors enabling AI assistants to connect to their data and services. And with agentic AI increasingly seen as the future of IT, MCP will only grow in use in the enterprise. But innovations like MCP also come with significant security risks . 3 industries where agentic AI is poised to make its mark July 4, 2024: IT leaders from finance, retail, and healthcare lend insights into what organizations are doing with AI agents today — and where they see the technology taking their organizations and industries in the future. IFS rolls TheLoops agentic AI into industrial ERP June 27, 2025: IFS is adding AI agent development and management capabilities to its ERP platform with the acquisition of software startup The acquisition brings TheLoops’ full Agent Development life cycle (ADLC) platform into IFS, enabling enterprises to design, test, deploy, monitor, and fine-tune AI agents with built-in support for versioning, compliance, and performance optimization. How AI agents and agentic AI differ from each other June 12, 2025 : With agentic AI in its infancy and organizations rushing to adopt AI agents, there seems to be confusion about the difference between “agentic AI” and “AI agents” technologies , but experts say there’s growing understanding that the two are separate, but related, tools. The future of RPA ties to AI agents June 10, 2025: RPA is accelerating toward a crossroads, with IT leaders and experts debating its future. Some IT leaders say that more powerful and autonomous AI agents will replace the two-decade-old AI precursor technology, while others predict that AI agents and RPA will work hand-in-hand. MCP is enabling agentic AI, but how secure is it? June 2, 2025 : Model context protocol (MCP) is becoming the plug-and-play standard for agentic AI apps to pull in data in real time from multiple sources. However, this also makes it more attractive for malicious actors looking to exploit weaknesses in how MCP has been deployed. The agentic AI assist Stanford University cancer care staff needed May 30, 2025 : At Microsoft Build 2025 earlier this month, Nigam Shah, CDO for Stanford Health Care, discussed agentic AI’s ability to redefine healthcare , especially in oncology, as physicians get overloaded with the administrative tasks of medicine, he said, which lead to burnout. Agentic AI, LLMs and standards big focus of Red Hat Summit May 26, 2025 : Red Hat, announced a number of improvements in its core enterprise Linux product, including better security, better support for containers, better support for edge devices. But the one topic that dominated the conversation was AI . Putting agentic AI to work in Firebase Studio May 21, 2025 : Putting agentic AI to work in software engineering can be done in a variety of ways. Some agents work independently of the developer’s environment, working essentially like a remote developer. Other agents directly within a developer’s own environment . Google’s Firebase Studio is an example of the latter, drawing on Google’s Gemini LLM o help developers prototype and build applications . Why is Microsoft offering to turn websites into AI apps with NLWeb? May 20. 2025 : NLWeb, short for Natural Language Web, is designed to help enterprises build a natural language interface for their websites using the model of their choice and data to answer user queries about the contents of the website. Microsoft hopes to stake its claim on the agentic web before rivals Google and Amazon do. Databricks to acquire open-source database startup Neon to build the next wave of AI agents May 14, 2025 : Agentic AI requires a new type of architecture because traditional workflows create gridlock, dragging down speed and performance. To get ahead in this next generation of app building, Databricks announced it will purchase Neon , an open-source serverless Postgres company. Agentic mesh: The future of enterprise agent ecosystems May 13, 2025 : Nvidia CEO Jensen Huang predicts we’ll soon see “a couple of hundred million digital agents” inside the enterprise. Microsoft CEO Satya Nadella takes it even further: “ Agents will replace all software .” Google to unveil AI agent for developers at I/O, expand Gemini integration May 13, 2025 : Google is expected to unveil a new AI agent aimed at helping software developers manage tasks across the coding lifecycle, including task execution and documentation. The tool has reportedly been demonstrated to employees and select external developers ahead of the company’s annual I/O conference. Nvidia, ServiceNow engineer open-source model to create AI agents May 6, 2025 : Nvidia and ServiceNow have created an AI model that can help companies create learning AI agents to automate corporate workloads. The open-source Apriel model, available generally in the second quarter on HuggingFace, will help create AI agents that can make decisions around IT, human resources and customer-service functions. How IT leaders use agentic AI for business workflows April 30, 2025 : Jay Upchurch, CIO at SAS, backs agentic AI to enhance sales, marketing, IT, and HR motions . “Agentic AI can make sales more effective by handling lead scoring, assisting with customer segmentation, and optimizing targeted outreach,” he says. Microsoft sees AI agents shaking up org charts, eliminating traditional functions April 28, 2025 : As companies increasingly automate work processes using agents, traditional functions such as finance, marketing, and engineering may fall away, giving rise to an ‘agent boss’ era of delegation and orchestration of myriad bots. Cisco automates AI-driven security across enterprise networks April 28, 2025 : Cisco announced a range of AI-driven security enhancements, including improved threat detection and response capabilities in Cisco XDR and Splunk Security, new AI agents , and integration between Cisco’s AI Defense platform and ServiceNow SecOps. Hype versus execution in agentic AI April 25, 2025: Agentic AI promises autonomous systems capable of reasoning, making decisions, and dynamically adapting to changing conditions. The allure lies in machines operating independently, free of human intervention, streamlining processes and enhancing efficiency at unprecedented scales. But David Linthicum writes, don’t be swept up by ambitious promises. Agents are here — but can you see what they’re doing? April 23, 2025: As the agentic AI models powering individual agents get smarter , the use cases for agentic AI systems get more ambitious — and the risks posed by these systems increase exponentially.A multicloud experiment in agentic AI: Lessons learned Agentic AI might soon get into cryptocurrency trading — what could possibly go wron April 15, 2025 : Agentic AI promises to simplify complex tasks such as crypto trading or managing digital assets by automating decisions, enhancing accessibility, and masking technical complexity. Agentic AI is both boon and bane for security pros April 15, 2025 : Cybersecurity is at a crossroads with agentic AI . It’s a powerful tool that can create reams of code in a blink of an eye, find and defuse threats, and be used so decisively and defensively. This has proved to be a huge force multiplier and productivity boon. But while powerful, agentic AI isn’t dependable, and that is the conundrum. AI agents vs. agentic AI: What do enterprises want? April 15, 2025 : Now that this AI agent story has morphed into “agentic AI,” it seems to have taken on the same big-cloud-AI flavor that enteriprise already rejected. What do they want from AI agents, why is “agentic” thinking wrong , and where is this all headed? A multicloud experiment in agentic AI: Lessons learned April 11, 2025 : Turns out you really can build a decentralized AI system that operates successfully across multiple public cloud providers. It’s both challenging and costly. Google adds open source framework for building agents to Vertex AI April 9, 2025: Google is adding a new open source framework for building agents to its AI and machine learning platform Vertex AI, along with other updates to help deploy and maintain these agents. The open source Agent Development Kit (ADK) will make it possible to build an AI agent in under 100 lines of Python code. It expects to add support for more languages later this year. Google’s Agent2Agent open protocol aims to connect disparate agents April 9, 2025 : Google has taken the covers off a new open protocol — Agent2Agent (A2A) — that aims to connect agents across disparate ecosystems .. At its annual Cloud Next conference, Google said that the A2A protocol will enable enterprises to adopt agents more readily as it bypasses the challenge of agents that are built on different vendor ecosystems not being able to communicate with each other. Riverbed bolsters AIOps platform with predictive and agentic AI April 8, 2025: Riverbed unveiled updates to its AIOps and observability platform that the company says will transform how IT organizations manage complex distributed infrastructure and data more efficiently. Expanded AI capabilities are aimed at making it easier to manage AIOps and enabling IT organizations to transition from reactive to predictive IT operations. Microsoft’s newest AI agents can detail how they reason March 26, 2025 : If you’re wondering how AI agents work , Microsoft’s new Copilot AI agents provide real-time answers on how data is being analyzed and sourced to reach results. The Researcher and Analyst agents take a deeper look at data sources such as email, chat or databases within an organization to produce research reports, analyze strategies, or convert raw information into meaningful data. Microsoft launches AI agents to automate cybersecurity amid rising threats March 26, 2025 : Microsoft has introduced a new set of AI agents for its Security Copilot platform, designed to automate key cybersecurity functions as organizations face increasingly complex and fast-moving digital threats. The new tools focus on tasks such as phishing detection, data protection, and identity management. How AI agents work March 24, 2025 : By leveraging technologies such as machine learning, natural language processing (NLP), and contextual understanding, AI agents can operate independently , even partnering with other agents to perform complex tasks. 5 top business use cases for AI agents March 19, 2025 : AI agents are poised to transform the enterprise , from automating mundane tasks to driving customer service and innovation. But having strong guardrails in place will be key to success. Nvidia launches AgentIQ toolkit to connect disparate AI agents March 21, 2025 : As enterprises look to adopt agents and agentic AI to boost the efficiency of their applications, Nvidia this week introduced a new open-source software library — AgentIQ toolkit — to help developers connect disparate agents and agent frameworks.. Deloitte unveils agentic AI platform March 18, 2025 : At Nvidia GTC 2025 in San Jose, Deloitte announced Zora AI , a new agentic AI platform that offers a portfolio of AI agents for finance, human capital, supply chain, procurement, sales and marketing, and customer service.The platform draws on Deloitte’s experience from its technology, risk, tax, and audit businesses, and is integrated with all major enterprise software platforms. The dawn of agentic AI: Are we ready for autonomous technology? March 15, 202 5: Much of the AI work prior has focused on large language models (LLMs) with a goal to give prompts to get knowledge out of the unstructured data. So it’s a question-and-answer process. Agentic AI goes beyond that. You can give it a task that might involve a complex set of steps that can change each time. How to know a business process is ripe for agentic AI March 11, 2025 : Deloitte predicts that in 2025, 25% of companies that use generative AI will launch agentic AI pilots or proofs of concept , growing to 50% in 2027. The firm says some agentic AI applications, in some industries and for some use cases, could see actual adoption into existing workflows this year. With new division, AWS bets big on agentic AI automation March 6, 2025 : Amazon Web Services customers can expect to hear a lot more about agentic AI from AWS in future with the news that the company is setting up a dedicated unit to promote the technology on its platform. How agentic AI makes decisions and solves problems March 6, 2025 : GenAI’s latest big step forward has been the arrival of autonomous AI agents . Agentic AI is based on AI-enabled applications capable of perceiving their environment, making decisions, and taking actions to achieve specific goals. CIOs are bullish on AI agents. IT employees? Not so much Feb. 4, 2025 : Most CIOs and CTOs are bullish on agentic AI , believing the emerging technology will soon become essential to their enterprises, but lower-level IT pros who will be tasked with implementing agents have serious doubts. The next AI wave — agents — should come with warning labels. Is now the right time to invest in them? Jan.13, 2025 : The next wave of artificial intelligence (AI) adoption is already under way, as AI agents — AI applications that can function independently and execute complex workflows with minimal or limited direct human oversight — are being rolled out across the tech industry. AI agents are unlike any technology ever Dec. 1, 2024 : The agents are coming, and they represent a fundamental shift in the role artificial intelligence plays in businesses, governments, and our lives. AI agents are coming to work — here’s what businesses need to know Nov. 21, 2024 : AI agents will soon be everywhere , automating complex business processes and taking care of mundane tasks for workers — at least that’s the claim of various software vendors that are quickly adding intelligent bots to a wide range of work apps. Agentic AI swarms are headed your way November 1, 2024 : OpenAI launched an experimental framework called Swarm. It’s a “lightweight” system for the development of agentic AI swarms, which are networks of autonomous AI agents able to work together to handle complex tasks without human intervention, according to OpenAI. Is now the right time to invest in implementing agentic AI? October 31, 2024 : While software vendors say their current agentic AI-based offerings are easy to implement, analysts say that’s far from the truth.
Apple’s most senior leaders so far seem to have failed to find consensus on how to move forward with Apple Intelligence . They recognize the company appears to be lagging in the generative AI (genAI) race, but management appears split between those who want to build it in Cupertino, and those who hope to buy their way to industry lead e rship . So far, neither side seems to have the room, with one thing holding everyone back being the insanely high valuations being thrown at AI companies at this point in the bubble. Why? Look it this way – the question has to be, “Why spend billions when some of these companies will be worth a great deal less once the AI bubble bursts , as it inevitably will?” It’s a reasonable position, given that valuations at the current levels are not sustainable . And while governments everywhere want to sell themselves cheap to climb into figurative bed with the current crop of genAI billionaires, they face massive public resistance to adopting the dystopian result of their dollar-drenched trysts. Russia’s intrusive Max app seems ripe to inspire similar behaviors from other authoritarian governments . Being different The other issue is differentiation. Apple is a product company , and for all the blather about artificial intelligence, the only thing that matters is how the tech can become part of its product family. If you glance at the many existing machine intelligence features already in its products, you’ll see that most of these supplement existing hardware. When it comes to Apple Intelligence, for Apple the North Star must be the need to ensure it continues to offer something unique. To some extent, doing this with AI is fundamentally difficult, as this kind of general purpose intelligence will eventually become a homogenous block of different models using similar data (all the data in the world) to inform responses to similar questions. With that inevitable homogenization, AI services may yet become utilities rather than differentiated products. So, it could make sense for Apple to avoid developing its own general purpose AI, resolving to create specific solutions that work much more effectively for specific use cases and relying on partnerships with these emerging AI-as-a-Service companies (AI-ASS?). That seems to be the Apple Intelligence way. In that picture, how much difference would an AI acquisition make? The problems of acquihire The only way in which such acquisitions might make a difference is if the company were to both gain access to the models and the people who made those models. If that’s how the company is looking at it, then any acquisition talks must necessarily require some commitments around employment and long-term loyalty to Cupertino. There’s no point doing an acquihire if everybody leaves, and this seems to be what’s been happening with the smaller AI purchases Apple has made . This may be an Apple-specific human resources problem, or a corporate culture problem, or it may reflect the insane competition for staff in the field . What this all boils down to is that if Apple can’t find an acquisition target that includes an employee transfer of people genuinely committed to Apple, then — other than IP and/or any political advantage it might gain — it doesn’t have a deal. It would just have less money in the bank and still be unable to find the talent. National security It’s worth noting that at least one of the current rumored Apple acquisition targets (Mistral), is likely to be seen as a company of national strategic importance to France. Mistral is France’s best-known AI company, with partnerships across the government and leading tech industry players. It’s unclear whether France, which, like the rest of Europe, is now struggling with data sovereignty, will see letting the AI service provider slip into the hands of US Big Tech fit the national narrative. What I’m saying is that as the national importance of AI is revealed, the number of potential acquisitions Apple can reasonably expect government approval for will decline. Made in California Can Apple invent its own AI to match the others on the market? The latest reporting suggests Craig Federighi, Apple senior vice president for software engineering still believes it is possible, with Services Senior Vice President Eddy Cue in the “Add to Shopping Basket” camp. And it is possible that much depends on what surprises Apple can bring to the table once it ships its own context-savvy Siri in a few months’ time. Earlier this month, Federighi told staff that his AI team has achieved more than was originally promised. “This has put us in a position to not just deliver what we announced, but to deliver a much bigger upgrade than we envisioned,” he said. If the company surprises and delights its audience, perhaps Apple will give those teams a little more time to build Apple’s own genAI solutions. If not, then perhaps those acquisition discussions will intensify. I can’t be certain but it is easy to imagine we’ll get a glimpse of some of what the company has put together during the iPhone launch on Sept. 9 . Partnerships Then there’s partnership possibility. The problem with partnerships is that teaming up gives people credibility, and to some extent gives the AI companies the upper hand. There are also unexpected challenges — Apple’s current AI partner, OpenAI, is reportedly building its own AI products with former Apple Chief Design Officer Jony Ive. We can’t know the extent to which that hardware threatens Apple’s interests. But it can’t have gone completely unnoticed that Apple is apparently about to introduce additional support for Google Gemini and other AI services as alternatives to the existing support Apple Intelligence has for ChatGPT . While the most recent chapter in Apple’s never-ending journey toward AI seems based on the narrative that the company can somehow buy its way to success, the truth is that any acquisition would be complicated. And in the absence of a determined management consensus, even with the company checkbook ready , no one seems to be making a move. All the same, as tides turn toward new iPhones, don’t be too surprised if something makes your awe drop. Follow me on social media! Join me on BlueSky , LinkedIn , and Mastodon .
Jacob Rajan and the vulture operated by puppeteer Jon Coddington in Paradise or the Impermanence of Ice Cream. Photos: supplied Discovering an ancient tradition under threat inspired Jacob Rajan and Justin Lewis’ play Paradise or the Impermanence of Ice Cream.
Rescue, by Pauline Bellamy. "Poems/Images" (Bellamys Metro) Dunedin is seeing an upturn in the number of new central city galleries, the newest of the lot being an urban gallery for Macandrew Bay’s Bellamys.
a2 Company co-directors Ben Ashby and Nadiyah Akbar. Photo: supplied New Zealand actor and director Ben Ashby and Australian dancer, choreographer Nadiyah Akbar are bringing their Melbourne Fringe-winning show Running into the Sun to Wānaka.
Anthropic has begun testing a Chrome extension for its Claude AI assistant with 1,000 paying subscribers, positioning it as a potential enterprise productivity tool but raising fresh concerns about security and control in the browser. The move brings Claude into direct competition with Microsoft’s Copilot in Edge and Google’s Gemini in Chrome, both of which are already weaving AI assistance into everyday workflows. But while Copilot and Gemini are primarily designed to help with tasks like summarizing and answering queries, Claude for Chrome takes things further. It can view the content a user is working on, take actions such as clicking buttons or filling out forms, and handle tasks directly within the browser. This highlights how the AI industry is moving beyond simple question-answering chatbots toward autonomous systems capable of completing complex, multi-step tasks across software applications. Anthropic’s rollout of Claude for Chrome is deliberately limited, framed as a debugging and security exercise rather than a full launch, underscoring how untested and fragile browser-based AI assistants remain. “Some vulnerabilities remain to be fixed before we can make Claude for Chrome generally available,” the company said in a statement. “Just as people encounter phishing attempts in their inboxes, browser-using AIs face prompt injection attacks—where malicious actors hide instructions in websites, emails, or documents to trick AIs into harmful actions without users’ knowledge.” Security concerns of AI browsers Analysts point out that the current browser privacy controls and plugins often provide an inadequate barrier, and AI extensions could deepen this vulnerability by collecting more data, potentially leading to a greater intrusion on user privacy and a weakening of overall security posture. This raises questions about whether the technology is mature enough to be considered for broader deployment. Unlike standalone AI chat apps, a browser-based assistant may have a direct line to corporate SaaS tools, potentially creating a far greater risk surface if exploited. “It’s critical to closely monitor and manage the use of these extensions,” said Neil Shah, VP for research at Counterpoint Research. “Any AI extension deployed in an enterprise environment must be enterprise-grade, task-specific, and governed by strict guardrails.” That risk is far from theoretical. As highlighted in recent research, large language models (LLMs) can be tricked with tactics as simple as run-on sentences, poor grammar, or manipulated images. “The answer isn’t to block innovation but to manage it by setting strict permissioning, sandboxing usage, and ensuring data policies are clear,” said Tulika Sheel, senior VP at market researcher Kadence International. “It’s about balancing productivity gains with a strong governance framework.” Enterprise adoption challenges ahead While Anthropic’s Claude for Chrome extension highlights growing interest in browser-based AI, third-party tools may struggle to gain traction in enterprise settings. “Native browser agents from Google (Gemini) and Microsoft (Copilot) are building deep, system-wide integrations into their ecosystems, making it challenging for competitors to offer a truly seamless and holistic experience,” Shah said. “Furthermore, a significant point of friction for third-party extensions is their limited ability to transfer learned data and context across different devices within tightly integrated ecosystems, such as Apple and Google.” This lack of interoperability could become one of the biggest barriers to broad enterprise adoption, leaving third-party agents at a disadvantage against vendors with tightly coupled platforms. “But adoption isn’t only about integration,” Sheel said, “It’s also about differentiation. If Claude can prove it delivers safer, more reliable, or more context-aware assistance, enterprises will take notice. Security concerns may slow uptake, but strong positioning around trust and usability can open doors.” For now, Claude for Chrome offers a glimpse of AI’s promise in the workplace and a reminder of the governance work ahead for CIOs.
US president warns progressive billionaire and allies that ‘we’re watching you’
New Zealander suffered multiple concussions Police called to his Nelson home on Wednesday morning Shane Christie, the former Māori All Blacks player who had wanted his brain to be studied after suffering from the effects of multiple concussions, has died aged 39. The New Zealand rugby player had campaigned for greater awareness in rugby of the impact of repeated blows to the head. After retiring from the game in 2017, Christie reportedly suffered from headaches, memory lapses, speech problems, depression and mood swings consistent with chronic traumatic encephalopathy (CTE). Continue reading...
Mitch Brown, the first AFL player to publicly identify as bisexual, has spoken about the fear of having his sexuality discovered while he was playing.
In the big-spending history of technology, there’s never been a time like this, when so many companies are spending so many hundreds of billions of dollars on a still-to-be proven technology — generative AI. Microsoft, Amazon, Meta and Google are expected to spend a combined $364 billion in their 2025 fiscal years alone. (Microsoft plans to spend $80 billion in AI capital expenditures .) That’s just the tip of the iceberg. It doesn’t take into account spending by chip manufacturers like Nvidia and Intel, or by genAI startups like Anthropic, OpenAI and all the rest. And it doesn’t take into account how much enterprises spend on the technology. Lost in all of this is an unthinkable thought: What if genAI can’t offer the benefits its backers claim ? What if it’s the biggest bust in a long line of tech busts? Microsoft, the largest and most valuable AI company in the world, has more to lose than any company if all the promises turn out to be illusory or overhyped. And there’s increasing evidence that might just be the case. From McKinsey to Gartner to MIT and beyond, analysts and researchers are saying the quiet part out loud: There may be more hype than reality to the genAI boom. Here’s what they’re saying — and what it could mean for the future of Microsoft if they’re right. Researchers: Most companies say using AI has been a bust Microsoft and other AI companies are bullish on the technology for a simple reason: Enterprises are betting big on it, spending tens of billions annually , and at an accelerating pace. IDC says enterprises will spend $61.9 billion this year on genAI alone , which doesn’t take into account all the other forms of AI. That $61.9 billion is nearly double the enterprise genAI spending from a year ago, the firm says. As enterprises are spending more , they’re finding that their investments aren’t paying off. A recent McKinsey report found that “nearly eight in 10 companies report using gen AI — yet just as many report no significant bottom-line impact. Think of it as the ‘genAI paradox.’” McKinsey isn’t alone. An MIT report, The GenAI Divide: State of AI in Business 2025 has even more disturbing news: 95% of genAI pilots in businesses are failing. A survey by the data and analytics firm S & P Global adds that 42% of companies abandoned most of their AI pilots by the end of 2024, up 17% from the previous year. Gartner has documented the booms and busts of new technologies for decades by tracking what it calls their “hype cycles.” The company has been tracking AI’s hype cycle and warns , “last year’s Hype Cycle for AI highlighted genAI as a potentially transformational technology with profound business impacts. This year, genAI enters the Trough of Disillusionment.” The researcher also found that, “Despite an average spend of $1.9 million on genAI initiatives in 2024, less than 30% of AI leaders report their CEOs are happy with AI investment return investment return. “ What this means for Microsoft Microsoft became a $3 trillion company based on its investments in AI. So, if businesses continue to find little value in AI projects, the company’s financial future will be endangered. But there’s a silver lining to the glum findings by McKinsey, S & P Global, Gartner and the rest. They all conclude the same thing: enterprises have been unhappy with their AI investments because they haven’t yet found out how best to use the technology. Once they do, AI spending can bring significant benefits. For example, the MIT study found that when companies turn to outside experts and firms for their projects, they succeed 67% of the time. But when they build them internally, they succeed only 33% of the time. McKinsey found that using AI agents specifically crafted to an enterprise’s workflow can offer tremendous benefits. It claims that the shift to agents “enables far more than efficiency. Agents supercharge operational agility and create new revenue opportunities.” And chief Gartner forecaster John-David Lovelock told The New York Times that when AI leaves behind the “Trough of Disillusionment” phase of its life cycle, it will become a highly useful productivity tool. Microsoft is well aware of all that. At its 2025 Build developers conference, execs proclaimed, “We’ve entered the era of AI agents” and made a host of announcements about its plans to help companies build agents using Copilot and its plans to build what it calls the “agentic web.” Microsoft also offers a guide to best practices for using Copilot in enterprises , including one titled, “How to deploy transformational enterprise-wide agents: Microsoft as Customer Zero.” The upshot None of the bad news about AI should be a surprise. It’s always been unclear at first whether a new technology will take off or fail. The current high rate of abandonment looks bad on the surface, but could be interpreted as a good sign — it shows companies are willing to experiment. It’s still too early to know whether genAI will crash and burn like the metaverse did — or transform business and the way we live as the Internet did. Microsoft’s fate depends on which way it goes.
State Highway One is closed near Waikouaiti in the South Island this evening after an oil tanker came off the road.
Detective Leading Senior Constable Neal Thompson. Photo: Supplied via Victoria Police A detective close to retirement and senior constable on temporary assignment have been identified as the officers shot dead in an alleged ambush.
Detective Leading Senior Constable Neal Thompson. Photo: Supplied via Victoria Police A detective close to retirement and senior constable on temporary assignment have been identified as the officers shot dead in an alleged ambush.
Otago Daily Times 
Computerworld NZ 
Financial Times 
The Guardian New Zealand 
SBS News 
RNZ