Responsible AI · Enterprise engineering · Global delivery
Curated analysis of AI, ML, and Agentic developments.
Feb 27, 2026
After talks over restrictions on mass surveillance and autonomous weapons collapsed, the US Defense Secretary announced on 27 February 2026 that Anthropic would be designated a supply chain risk; the company vowed to sue.
Feb 21, 2026
The India AI Impact Summit in New Delhi closed on 21 February 2026 with a declaration endorsed by 92 countries and bodies and more than 200 billion dollars in announced AI investment commitments.
Feb 09, 2026
Bitsight found more than 30,000 internet-exposed OpenClaw agent instances between 27 January and 8 February 2026, following a one-click RCE flaw, a platform data leak and hundreds of malicious add-on skills.
Feb 05, 2026
On 5 February 2026 Anthropic released Claude Opus 4.6 with agent teams and a 1M-token context, hours after OpenAI launched GPT-5.3-Codex, its first model rated 'High' for cybersecurity capability.
Jan 27, 2026
Moonshot AI released Kimi K2.5 on 27 January 2026: a one-trillion-parameter open-weight mixture-of-experts model with vision and an agent-swarm mode that can run up to 100 sub-agents.
Jan 22, 2026
On 22 January 2026 Anthropic released a rewritten constitution for its Claude models, ordering priorities as safe, ethical, guideline-compliant and helpful, and licensing the text under CC0.
Jan 15, 2026
California's frontier-AI transparency law and Texas's AI governance act took effect on 1 January 2026, even as Executive Order 14365 set 90-day deadlines for federal action against 'onerous' state rules.
Jan 07, 2026
OpenAI introduced ChatGPT Health on 7 January 2026, a separate space that connects medical records and wellness apps with isolated storage, no training on health chats, and explicit non-diagnostic limits.
Jan 05, 2026
Nvidia announced its six-chip Vera Rubin platform on 5 January 2026, claiming up to a tenfold cut in inference token cost versus Blackwell, with partner availability in the second half of 2026.
