Hardware
Anker unveils Eufy MindBase hub with on-device LLM for home security
Device processes camera video without sending data to the cloud, using an AI model trained by Anker itself.
AI on the line: security, data, and the future of local processing
This week brought AI investments in home security, advanced cybersecurity, and kernel Linux vulnerability scanning. Amid the race for more powerful models and the debate over data privacy, the question remains: where to draw the line between automation and control?
IA
OpenAI nears launch of cybersecurity-focused AI model
The company is developing a persistent AI agent with security capabilities and adopting a stricter isolation posture after a data theft incident.
From the studio
AI on the line: security, data, and the future of local processing
This week's edition lays bare a dilemma spanning hardware, software, and services: who actually controls the data we produce and the systems that protect us? Anker bets on local AI processing, OpenAI trains security agents — but the Linux kernel already logs 2,000 vulnerabilities per version and maintainers are calling for help.
While Fujitsu stacks caches on separate dies for performance gains, Anthropic unifies memories between Claude and Cowork, centralizing user profiles. The promise of convenience runs into the reality that every new layer of automation is also a new attack surface.
The week's rule of thumb: before adopting any AI solution, ask who holds the key to the safe. If the answer is a third party, the risk is yours.
In this issue
Segurança
Linux kernel hits 2,000 CVEs per version with AI scans
Kernel maintenance reaches 2,000 registered vulnerabilities per release as maintainers say they are overwhelmed.
Hardware
Fujitsu Monaka stacks full cache on separate 5nm die
Fujitsu's Japanese processor stacks all cache on a dedicated die, arriving in 350W and 500W versions with samples available and volume production planned for 2027.
Segurança
OpenClaw 2.0 eases install but keeps security risks on users
New version of the OpenClaw AI agent harness prioritizes ease of use and interface, but the project foundation says security responsibility remains with the user.
Dados
Claude and Cowork now share a single memory system
Anthropic has unified memory between Claude and its Cowork platform, creating a single profile that moves between both work environments.
Sections