Senior Software Engineer
Browse published Educatifu articles attributed to Michael Carter.
Researchers documented the first ransomware operation run end-to-end by an AI agent. It used no zero-days — just exposed, unpatched infrastructure. Here's what to do about it.
Microsoft and AWS have committed billions to hands-on enterprise AI deployment teams. The pivot from selling models to selling outcomes tells you where the real bottleneck is.
Humanoid robots are moving from lab demos to factory floors and public IPOs. A grounded look at where physical AI actually works today — and where the hype outruns reality.
AI data centres don't just consume electricity — they consume water, at scale. A look at the cooling crunch, the startups racing to fix it, and why it matters to your cloud costs.
Open-weight Chinese models are 60–90% cheaper than US frontier models and increasingly competitive. Here's how to weigh cost against governance, security and data sovereignty.
AI data-centre demand is driving memory prices up sharply and power consumption to record highs. What the infrastructure squeeze means for ordinary IT budgets — and how to plan.
Google just invested in nuclear fusion, joining a rush by hyperscalers to secure their own electricity. Here's why AI's energy hunger is reshaping how data centres get powered.
The UN launched a global AI governance commission and the US is negotiating voluntary model-release standards. Cutting through the noise — what actually affects EU firms.
AI assistants now write a large share of production code — and the tools themselves have become an attack surface. Concrete controls for adopting them safely.
38% of organisations are piloting AI agents, yet only 11% run them in production — and Gartner expects 40% of agentic projects to be cancelled. How to be in the other 60%.
DORA enforcement is now active and Poland's NIS2 law took effect in April 2026. Who is in scope, what each regime demands, and a pragmatic 90-day plan.
How AI-based fraud detection works, what a production-grade pipeline looks like, and the pitfalls fintech teams should avoid — a practical 2026 guide.
Worldline, ING and Visa complete a live agentic payment in Germany, Visa launches a threat intelligence platform, and Plaid weighs a US IPO.
CISA flags active exploitation of SharePoint CVE-2026-45659, a massive password-spray campaign hits Azure CLI, and AI coding tools face zero-click prompt injection.
OpenAI previews GPT-5.6 Sol, IBM and Red Hat commit $5B to fixing AI-discovered open-source bugs, and researchers expose agentic browser risks.
Executive orders target a commercially relevant quantum computer by 2028 and pull the US post-quantum deadline forward to 2031 — plus QuantumEAGLe and more.
Quantum computers will eventually break today's public-key encryption, and "harvest now, decrypt later" makes it a present-tense risk. A no-panic roadmap — inventory, prioritise, migrate.
AI assistants have moved from autocomplete to agents that draft whole features. Here's what actually changes for a development team — and what doesn't — plus a practical adoption guide for smaller companies.
Seven practical questions for assessing delivery, security, commercial and handover risk before signing an IT services contract.
Moving to the cloud doesn't have to be risky. This step-by-step checklist covers assessment, planning, execution and the week after go-live.
Before buying another security product, review the fundamentals. These five control areas provide a practical starting point for reducing avoidable security risk.
Learn how Educatifu handles attribution, review labels and corrections in our editorial policy.