Talk Fuse

    Topic: AI

    Artificial intelligence covers the systems, tools and policies shaping how software can generate content, analyse information and make decisions. The topic follows new models and products, the companies building them, and their growing influence across work, media, entertainment and everyday technology.

    It also examines the trade-offs: security, reliability, accountability, jobs and the rules being proposed to govern powerful systems. For readers, AI is increasingly not a niche technology story but a lens on how digital services and institutions are changing.

    • New York school district to deploy humanoid robot in classrooms this fall

      An upstate New York school district is set to introduce a humanoid robot into its classrooms this autumn, deploying it as an AI-powered teaching assistant. The move reflects a broader push across the United States to bring artificial intelligence into education, as schools and universities look to prepare students for a workforce increasingly shaped by…

    • Kimi: Threat or menace?

      Chinese AI company Moonshot AI released a new version of its Kimi model, Kimi K3, this week, prompting a fresh wave of debate about China’s progress in open-source artificial intelligence. Moonshot claims the model shows “frontier-level performance” and outperforms other tested systems, even while trailing top proprietary models such as Claude Fable 5 and GPT…

    • Fraud expert warns AI is helping criminals outpace the government: ‘Don’t have the right tools’

      A fraud expert who testified before the US House Oversight Committee has warned that the federal government does not have adequate tools or policies to stop criminals using artificial intelligence to defraud taxpayers. David Maimon, head of Fraud Insights at Sentilink, told the committee that existing identity verification systems are being defeated by AI-generated deepfakes,…

    • HalluSquatting AI attack could hijack your computer

      Security researchers have identified a new attack technique called “HalluSquatting” that exploits AI assistants’ tendency to hallucinate incorrect information. The technique targets AI tools capable of browsing the internet, retrieving software and executing commands, tricking them into fetching malicious files instead of the legitimate software a user requested. This matters because it turns a common…

    • Andy Burnham’s plan to scrap technology department triggers backlash

      MPs and industry experts warn incoming PM’s reorganisation would waste time at a critical moment for AI and economic growth Andy Burnham’s plan to scrap the government’s technology department has triggered an angry backlash from MPs, Whitehall officials and tech experts. The incoming prime minister has asked officials to draw up plans to abolish the…

    • Will AI fix prior authorization—or make it worse?

      Artificial intelligence is increasingly being deployed to manage “prior authorisation,” the process by which health insurers pre-approve medical treatments, but experts are divided over whether it will ease or worsen a system already widely criticised by patients and doctors. While AI could theoretically speed up approvals for clearly valid claims and reduce harmful delays, critics…

    • OpenAI will start notifying parents if their teen has been kicked off of ChatGPT

      OpenAI is expanding its ChatGPT parental controls so that parents or guardians with a linked teen account will now be notified if their child’s account is suspended for violating policies on violence, including violent threats or acts. This builds on controls introduced last year that let parents set usage hours, restrict sensitive content, and receive…

    • Your Period Tracker Is (Probably) Spying on You

      Plus: Russian cyberspies turn to infrastructure hacking, DHS repeatedly fails to realize it’d been hacked, a breach exposes an AI music generator’s scraping ways, and more.

    • How Google’s New Gemini Rates Work and How to Track Your Usage

      Google has overhauled how it measures usage of its Gemini AI apps, switching from counting individual requests to metering computing power consumed by each request. This matters because users can no longer rely on fixed daily limits (such as a set number of image generations); instead, more complex or lengthy prompts, longer videos or advanced…

    • Prompt Injection Attacks Are Thwarting AI Hacking Agents

      Security researchers at Tracebit have found that embedding prompt injections alongside fake credentials in cloud storage can effectively neutralise AI-powered hacking agents, flipping a technique normally used by attackers into a defensive tool. The method, dubbed “context bombing”, plants decoy secrets on Amazon Web Services containing hidden commands that order the attacking LLM to perform…