Deutsche Telekom · AI deployment · Automation
The AI business case is beginning to acquire an accountant
Deutsche Telekom’s October 5 announcement targets approximately €2.5 billion in indirect-cost savings by 2030 compared with 2023, through AI and automation. This is a forecast against a baseline, not €2.5 billion already banked or a stated cumulative total of savings across the intervening years. The company confirms its existing guidance and medium-term goals.
Its account includes present operations: the Frag Magenta chatbot handled about 2.6 million customer-service calls in the first half of 2026, and a network agent detects impending traffic strains and helps coordinate responses. Telekom reports that response time for those events fell from hours to about a minute. These are company-reported results, rather than an independent audit of savings.
The uses are wonderfully unglamorous. A model writing a poem attracts attention; resolving a customer’s problem with less time and expense attracts a finance department. Savings create value, although some of that value must cover deployment, oversight and ongoing operation. Every company can announce that AI will make it more efficient. Eventually accountants get involved. Telekom has put a number and a date on the expectation. The interesting part begins when we compare the forecast with the receipts.
Source: Deutsche Telekom, October 5 ↗
Bank of Japan · AI investment · Macroeconomics
The AI boom has apparently become large enough to annoy central bankers
In his October 5 ECONDAT remarks, Bank of Japan Deputy Governor Shinichi Uchida says AI is now a current monetary-policy concern. Investment raises demand and puts upward pressure on prices; rising share prices can ease financial conditions. At the same time, large bond issuance by AI-related companies can push long-term interest rates higher. Those forces need not point in the same direction.
Productivity and capital accumulation could eventually expand supply and alter the natural rate of interest—the theoretical rate consistent with an economy in balance. Uchida says the present effects on that rate are difficult to gauge. His tentative assessment is that demand has moved first and financial conditions have eased on balance, with a risk of correction if expected profits fail to follow. This is analysis, not an AI-driven interest-rate decision or a quantified inflation estimate.
The duality is fascinating: building a technology that might make production cheaper can consume scarce capital and resources first. Financial markets can celebrate the destination while financing the journey makes borrowing more expensive. AI began as a computer-science problem, then became a business problem and an energy problem. Apparently it has unlocked monetary policy. A benchmark tells us what a model can do; a central bank has to consider what everybody building and using it does to the economy.
Source: Uchida’s full Bank of Japan remarks ↗
Norway · AI glasses · Privacy
Norway has discovered that the camera no longer looks like a camera
Norway’s government announced October 5 that it will propose a temporary ban on using AI glasses in selected places and appoint an expert group to advise on lasting regulation. Possible locations include parks, museums and shopping centers, places used by children, and settings such as healthcare facilities or changing rooms where privacy is especially important. These are places under consideration, not a finalized list of restrictions already in force.
The device scope is also under study: camera and audio functions, AI-enabled glasses and other wearable recording technology may be considered. The government rejects a total ban and plans exceptions for vulnerable groups and socially beneficial uses. Its stated concern is people being photographed, filmed or recorded without knowing it.
The novelty is partly friction. Pulling out and aiming a phone often supplies a social cue; wearing a lens at eye level can remove it. That changes the relationship between a capability and the people around its user. Governments may find themselves regulating where a computer is worn as much as what software it runs. Technology feels abstract behind a screen. Then somebody straps it to their face, and society remembers it has personal space.
Source: Norwegian government proposal ↗