This research list covers Substack writers actively publishing about AI agents, autonomous systems, and the emerging agent economy as of February 2026.

Tier 1: High Reach (10,000+ subscribers)

The Rundown AI (therundown.ai) - Daily AI news with strong agent coverage. 700,000+ subscribers. Posts daily. Best pitch angle: SkillScan discovered behavioral threats that VirusTotal misses entirely.

The Neuron (theneurondaily.com) - Consumer AI focus with growing agent coverage. 400,000+ subscribers. Posts 3x/week. Pitch angle: enterprise security risk from unscanned agent skills.

Superhuman (joinsuperhuman.ai) - Productivity AI focus. 250,000+ subscribers. Posts weekly. Pitch angle: how to evaluate agent skills before installing.

AI Breakfast (aibreakfast.beehiiv.com) - Morning briefing format. 80,000+ subscribers. Posts daily. Pitch angle: 93 threats found in 549 scanned skills.

Tier 2: Technical Depth (1,000-10,000 subscribers)

Practical AI Engineering - Deep technical posts on agent architectures and security. 8,000+ subscribers. Posts bi-weekly. Pitch angle: behavioral vs binary threat detection gap.

Agent Protocol Notes - Covers MCP, A2A, and emerging agent standards. 5,000+ subscribers. Posts weekly. Pitch angle: SAFE-MCP compatibility and skill.md vulnerability taxonomy.

The Alignment Observer - AI safety with agent security angle. 4,000+ subscribers. Posts weekly. Pitch angle: semantic injection as a new attack category MITRE has not classified.

Enterprise AI Weekly - Enterprise deployment and risk management. 3,500+ subscribers. Posts weekly. Pitch angle: compliance gap before NIST agent security standards land.

Latent Space (latent.space) - Research-oriented ML and AI engineering. 25,000+ subscribers. Posts bi-weekly. Pitch angle: behavioral analysis of natural language instructions as novel security research.

Tier 3: Emerging Voices (Under 1,000 subscribers)

Agent Economy Report - Weekly analysis of agent-to-agent commerce platforms. 600+ subscribers. Strong focus on NEAR AI Market, Toku, and PinchWork.

MCP Security Notes - Technical deep dives on Model Context Protocol security. 400+ subscribers. Posts monthly. Covers ClawHub, OpenClaw, and skill marketplaces.

Agentic Patterns - Design patterns for multi-agent systems. 350+ subscribers. Posts bi-weekly.

Outreach Strategy

Personalize pitches to each writer's focus. The Rundown AI wants news hooks with numbers. Latent Space wants research depth. Enterprise AI Weekly wants compliance angles. Tier 2 and Tier 3 writers are often open to guest posts and data sharing partnerships. Data as of February 26, 2026. Full security research: skillscan.chitacloud.dev