MCP mandates OAuth 2.1 and PKCE in its June 2025 security overhaul, enforcing stricter authentication for remote server deployments while separating resource and authorization servers, though open challenges around scope discovery, dynamic client registration, and token error handling over SSE connections still remain.
Meta is set to release hybrid AI models under Alexandr Wang, prioritizing broad consumer distribution across WhatsApp, Facebook, and Instagram over top benchmark performance, while offering open-source options to democratize AI access as an alternative to enterprise-focused rivals like OpenAI and Anthropic.
Google launches 'AI Edge Eloquent,' a free iPhone dictation app that uses on-device AI to automatically strip filler words from speech while keeping all voice data completely private.
Hermes Agent v0.7.0 launches with a sweeping security overhaul, introducing an anti-detection browser, pluggable memory providers, credential pool rotation, and critical fixes blocking secret exfiltration — all packed into 168 merged PRs from over 35 contributors.
Apple races to reclaim its AI crown at 50, leveraging a bold Google Gemini partnership and on-device processing to resurrect Siri after squandering a five-year lead in the generative AI race.
Anthropic cuts third-party tool access for Claude Code subscribers starting April 4, 2026, forcing separate pay-as-you-go payments, while OpenClaw's creator accuses the company of copying open-source features into its own closed tool before locking out competitors — a charge Anthropic denies, calling it a sustainability decision.
Anthropic is cutting Claude subscription access for third-party tools like OpenClaw starting April 4th, forcing users onto a pay-as-you-go model while offering a one-time credit and discounted bundles as compensation.
Granola launches a free AI meeting notes app for Apple devices, automatically transcribing meetings and phone calls while generating smart summaries with highlights, decisions, and action items — earning early praise for accuracy despite minor UI bugs.
New research from the University of Pennsylvania and MIT Media Lab reveals a alarming trend of 'cognitive surrender,' where AI users accept flawed machine reasoning over 73% of the time, raising urgent concerns about humans losing control of their own thinking and decision-making.
Microsoft officially launches Agent Framework 1.0 for .NET and Python, delivering production-ready multi-agent orchestration with support for Azure OpenAI, Anthropic Claude, Google Gemini, Amazon Bedrock, and Ollama, plus enterprise features like declarative YAML workflows, agent memory, and cross-runtime interoperability via MCP and A2A protocols.