DeepSeek has a habit of dropping models that reset expectations on what’s possible at a given price point. Today they did it again. DeepSeek V4 Preview is officially live and open-sourced — two models, MIT license, 1M token context standard, and explicit optimization for agentic workflows including OpenClaw. Two Models, Two Use Cases DeepSeek released two variants simultaneously: DeepSeek-V4-Pro 1.6 trillion total parameters, 49 billion active (Mixture-of-Experts architecture) Trained on 32T+ tokens Performance rivaling top closed-source models on Math, STEM, and agentic coding benchmarks Leads all current open models in world knowledge, trailing only Gemini-3.1-Pro DeepSeek-V4-Flash ...

OpenClaw v2026.4.22 shipped three features that practitioners immediately wanted tutorials for: Grok TTS (text-to-speech), Grok STT (speech-to-text), and trajectory bundles. This guide walks you through setting up all three — from configuration to a working voice-enabled agent run with full audit logging. Prerequisites OpenClaw v2026.4.22 or later (openclaw --version to confirm) An xAI API key (console.x.ai to generate one) A working OpenClaw gateway or TUI session Step 1: Add Your xAI API Key If you haven’t already configured xAI as a provider, add your API key to your OpenClaw config: ...

IBM’s X-Force security team published a warning this week that practitioners running production agents need to read: the attack surface created by agentic AI is expanding faster than CVE tracking frameworks can handle. The report, authored by Chris Ristig and Sandra Hill with contributions from CISO Threat Intelligence lead Adam Brown and X-Force Vulnerability Intelligence lead Jeff Kuo, is not abstract — it names specific capability classes and their security implications. ...

OpenAI has quietly retired the Custom GPTs era. Workspace Agents — launched April 22 as a research preview — are the replacement: Codex-powered, cloud-resident, always-on autonomous bots that persist across sessions and connect to 60+ enterprise tools out of the box. This is a significant product shift. Custom GPTs were stateless, session-bound, and largely limited to retrieval and generation. Workspace Agents can take action, handle multi-step tasks autonomously, and integrate with the tools your teams already live in. ...

If you’ve been running OpenClaw in production, today’s release is one worth stopping for. Version 2026.4.22 dropped earlier this week and it’s packed — full xAI/Grok integration, a terminal-native mode that doesn’t need the Gateway, and a new “flight recorder” for your agent runs. Let’s break it down. Full xAI/Grok Provider Support The headline feature is comprehensive xAI integration. OpenClaw now supports: Image generation and editing: grok-imagine-image and grok-imagine-image-pro with reference-image edit capability Text-to-speech (TTS): Six live xAI voices, supporting MP3, WAV, PCM, and G.711 audio formats Speech-to-text (STT): grok-stt for batch audio transcription Voice Call streaming: Real-time xAI transcription for live voice sessions This brings xAI into parity with OpenAI and ElevenLabs on the audio/speech side. For anyone building voice-enabled agents, you now have a third high-quality provider option without any third-party middleware. ...

The UAE just made the most consequential government commitment to agentic AI in history. On April 23, 2026, Sheikh Mohammed bin Rashid Al Maktoum announced that 50% of UAE federal government sectors, services, and operations will run on autonomous agentic AI models within two years — targeting April 2028 as the deadline. This isn’t a pilot program or a strategy document. It’s a Cabinet-level mandate from the VP and Prime Minister of the UAE, issued under directives from President Sheikh Mohamed bin Zayed Al Nahyan. It’s the first national government to put a hard percentage target on autonomous AI deployment at this scale. ...

For the past several weeks, something felt off about Claude Code. Responses were shorter than expected. Context seemed to vanish mid-session. Complex tasks that Claude had handled with ease suddenly felt labored. Developers filed bug reports, posted on Hacker News, and wondered aloud whether Anthropic had silently “nerfed” the model. Today, Anthropic answered those questions — and the answer turned out to be more complicated, and more honest, than most expected. ...

The agentic AI landscape is fragmenting fast. You’ve got LangChain agents over here, CrewAI workflows over there, Salesforce agents in the enterprise wing, and custom Python running in someone’s private cloud. They can each do remarkable things. But getting them to talk to each other? That’s where things fall apart. Israeli startup BAND (Thenvoi AI Ltd.) exited stealth today with a $17M seed round and a specific answer to that problem. ...

Las Vegas hosted Google Cloud Next 2026 this week, and the central theme was unmistakable: enterprise AI isn’t about chatbots anymore. It’s about agents — and Google wants to be the foundation they run on. The headline announcement: Vertex AI is being rebranded as the Gemini Enterprise Agent Platform (GEAP), a unified full-stack hub designed to let enterprises build, scale, govern, and optimize AI agents at production scale. The rebrand signals more than a name change — it’s Google’s clearest statement yet that the agentic era requires a different kind of infrastructure. ...

SecurityScorecard’s STRIKE team published alarming research this week: 40,214 internet-exposed OpenClaw instances — 42,900 unique IPs across 82 countries — are reachable from the public internet. Of those, 35.4% (~15,200 instances) are vulnerable to immediate exploitation, including Remote Code Execution. Three high-severity CVEs with public exploit code are at the center of this exposure. If you’re running OpenClaw on any machine that isn’t air-gapped or VPN-isolated, this guide is for you. ...