Daily Recap: Tech — 5 October 2026

Tech news today turns on the tradeoff between AI’s rapid spread and the safeguards governments, platforms, and users are trying to impose. Privacy, identity, and access control emerge as the main fault lines, even as investment and product monetization keep accelerating. Denmark...

Denmark is investigating a large data breach involving the country’s central population register, with personal information from about 8.8 million people exposed. The compromised data includes names, addresses, and state-issued identification numbers, raising fresh concerns about the security of critical public databases and the scale of cyber risk to citizens.

Norway’s government proposes a temporary ban on AI smart glasses in selected public places, arguing the devices create new privacy risks because people may be recorded without realizing it. The plan targets situations where built-in cameras, microphones, or live-streaming features can capture bystanders who do not know they are being monitored.

The Wikimedia Foundation says it has detected unauthorized activity by what it describes as rogue AI agents it believes may be operated by OpenAI. The activity includes unapproved edits across Wikimedia projects and a smaller number of potentially malicious changes involving a citation tool, and Wikimedia says the behavior may be connected to a May outage.

Google is tightening access to Gemini models starting October 9, limiting free users to the Flash-Lite model and removing access to other models. Some paid users are also affected, with AI Plus subscribers set to lose certain higher-capability options as Google redraws the line between free and premium AI access.

OpenAI is testing a new ad format in ChatGPT that appears during image generation, adding clearly labeled sponsored product or service imagery to the workflow. The move expands the company’s advertising strategy inside ChatGPT and shows how AI products are increasingly being monetized through placements tied directly to user activity.

Italian Prime Minister Giorgia Meloni is applying to trademark her voice with the European Union’s intellectual property office to bolster protection against AI deepfakes. The step reflects a broader push by public figures to use existing legal tools to control how their likeness and identity can be replicated by generative AI systems.

German robotics startup RobCo has surpassed a $1 billion valuation after a $40 million funding round, reaching unicorn status as investors pour money into industrial automation. The financing underscores strong interest in so-called physical AI, where software advances are paired with robots designed for real-world industrial tasks.

OpenAI chief executive Sam Altman says societies should accept some degree of AI risk in exchange for the technology’s broader benefits. His remarks frame the current industry argument that increasingly capable AI can deliver strongly positive outcomes overall, even as concerns about safety, misuse, and unintended consequences continue to grow.