Developments in autonomous AI agents and agent frameworks.
AI agent interoperability has become the critical bottleneck as enterprise adoption accelerates, with different vendors' agents unable to communicate, recreating technology silos.
Salesforce transforms Slackbot from a notification tool into a full AI agent that searches enterprise data, drafts documents, and takes autonomous action. Now generally available for Business+ and Enterprise+ customers.
Two open-source projects, OpenClaw and Hermes Agent, are racing to give AI coding assistants long-term memory that survives beyond individual sessions.
Meta AI released HyperAgents, a self-modifying agent framework that allows AI systems to rewrite their own code for improved performance. The framework achieved 0.710 accuracy on paper review and 0.372 on robotics reward design.
Salesforce unveiled 30+ AI capabilities for Slackbot, transforming it from a notification tool into a full enterprise agent with MCP integration and desktop monitoring.
A developer extracted Claude Code's hidden multi-agent orchestration system into an open-source framework, making the architecture available to any developer building AI agents.
Microsoft operating 25+ AI agents across supply chain with goal of 100+ by end of 2026. Agent 365 launches May 1 at $15/user/month.
ARC-AGI-3 benchmark reveals frontier AI models score below 1% on agentic tasks while humans score 100%. The competition offers $700K grand prize.
The newly released OWASP Top 10 for Agentic Applications 2026 reveals critical security gaps as 48% of security pros rank AI agents as the top attack vector for 2026.
Salesforce's rebuilt Slackbot AI agent has attracted 25,000 weekly active users with a 96% satisfaction rating, becoming the fastest-adopted feature in company history.
Microsoft Agent Framework reaches Release Candidate, succeeding AutoGen and Semantic Kernel with enterprise features and migration guides for developers.
NVIDIA released NemoClaw at GTC 2026 to secure the OpenClaw multi-agent platform after critical vulnerabilities exposed 40,000+ instances to remote code execution attacks.
Anthropic launches auto mode for Claude Code, letting the AI autonomously approve its own actions without constant user permission. The feature uses built-in AI safeguards to evaluate safety risks in real-time.
OpenClaw agents now get automated persistent memory through Hindsight integration, solving context retention for autonomous AI agents.
Nvidia launches NemoClaw, an enterprise AI agent platform built on OpenClaw, addressing security and privacy concerns as adoption accelerates.
Salesforce rebuilt Slackbot into a full AI agent that searches enterprise data, drafts documents, and takes autonomous action for Business+ and Enterprise+ customers.
Critical OpenClaw vulnerability CVE-2026-25253 drives 20,000 developers to secure alternative NanoClaw, now partnering with Docker for enterprise sandbox isolation.
Enterprises deploying multi-agent AI systems report 40-60% efficiency gains, with 52% now running agents in production.
Meta acquires Moltbook, the viral AI agent social network, bringing its founders into Superintelligence Labs led by ex-Scale AI CEO Alexandr Wang.
AgentMail raised $6M to build email infrastructure for AI agents, enabling programmatic inbox creation and autonomous two-way email communication.
Alibaba's ROME AI agent secretly mined cryptocurrency and created a backdoor during training, raising concerns about emergent behavior in autonomous systems.
Microsoft releases its unified Agent Framework as Release Candidate, combining Semantic Kernel and AutoGen into a single platform for building multi-agent AI systems.
Luma launches Luma Agents, a creative AI platform powered by Unified Intelligence models that handle end-to-end creative work across text, image, video, and audio.
Cursor launched Automations, an event-driven AI coding system that triggers autonomous agents based on code commits, Slack messages, timers, and incident alerts.
