Japan · Emergency response · AI-assisted defense
Japan sets aside ¥10 billion for urgent cyber defenses
Japan announced an emergency cybersecurity package on November 6 after a series of breaches affecting companies and government bodies. The government set aside about ¥10 billion, roughly $63 million, from budget reserves for immediate measures. That is the initial funding for the response.
The Ministry of Economy, Trade and Industry plans to put a specialist team in place by the end of November to use AI in vulnerability testing. The package also calls for sharing threat information across government and industry and coordinating rapid help for organizations hit by attacks.
Finding a flaw gives defenders a place to start. Someone still has to fix it, check whether attackers used it and help the affected organization restore service. Japan is putting money into each part of that job. The announcement leaves the methods behind the earlier breaches open.
Japan’s November 6 announcement ↗Reuters on the initial funding ↗Amsterdam · Data centers · Public opposition
Protesters enter a Microsoft data center construction site
On Friday, more than 100 people gathered at a data center construction site in Amsterdam. Around 100 made it onto the site, according to a Reuters witness. Some climbed cranes. Pure Data Centres owns the site, which is being developed for Microsoft.
The action came at the start of two days of coordinated protests across Europe. Campaigners want a pause on large new projects and cite pressure on local electricity, water and land. Microsoft said it respected peaceful protest and opposed unlawful actions. Pure Data Centres said it remained committed to engaging with local stakeholders.
The dispute puts a local question next to the global demand forecasts. A developer can show how much computing capacity it plans to sell. The people living near the site want to know how much power and water it will need, where those supplies will come from and who pays for upgrades.
Reuters on the protests and company responses ↗Reuters’ on-site account ↗Cybersecurity · Refrigeration controls · Espionage
What the freezer attacks tell us, and where AI actually showed up
Friday’s Axios follow-up put the European refrigeration attacks alongside a separate set of intrusions with documented AI use. In the refrigerator cases, attackers altered temperature settings remotely. That is enough to spoil stock when a control system governs equipment that stores food. Whether AI played any part remains unverified.
CrowdStrike documented a different campaign in which a group it calls VAPOR PANDA used the ARTEX framework with DeepSeek and a second commercial model. The company observed activity directed at organizations in at least 14 countries over 11 days and confirmed follow-on activity after gaining access at nine organizations. Recovered transcripts showed a human operator steering agents as they worked on multiple targets in parallel.
The cases pose distinct problems for defenders. A remotely exposed controller can already put a physical process within reach of an intruder. In the espionage campaign, one operator could direct simultaneous engagements. That gives defenders less time to catch an intrusion before it spreads.
Axios’ November 6 follow-up ↗Claroty’s analysis of refrigeration exposure ↗CrowdStrike’s campaign investigation ↗