Accel Partner Event

Attendees: Dustin J Ross Date: April 8, 2026 Type: External

Summary

AI Security Market Opportunity

• Legacy cybersecurity tools completely unprepared for AI context
• Net new generation of AI tools releasing faster than security can keep up
• Opportunity for entrepreneurs to rebuild entire security stack
• AI security market potentially larger than current AI market
  • Agents will need security agents around them
  • Focus on pragmatic solutions vs doom scenarios

Enterprise CISO Challenges

• Cannot be seen as “office of no” while everyone pressured to adopt AI
• PTSD from SaaS adoption where they initially resisted, then had cleanup problems
• Most thoughtful approach: lean in but choose internal champions as partners
• Some putting AI agents in HR systems for centralized control/shutoff capability
• Shadow IT for AI tools inevitable but bulk adoption will consolidate

Cybersecurity Sales Environment

• CISOs extremely difficult to reach - most “wined and dined” people on planet
• Industry driven by sales vs R&D (visible in investment ratios)
• 50+ similar solutions using same messaging at conferences
• Breakthrough strategy: find differentiated technology + strong reference customers
• Early reference customers critical - others follow when they see adoption

Depth First Technology Advantage

• Interdisciplinary team: CTO from DeepMind (Gemini reinforcement learning)
• Combines cutting-edge AI expertise with cyber security operators at scale
• Performance proven on cyber gym benchmarks
• Compounding learning system advantage with Andre leading
• Anthropic building similar internal tools, validating market need

AI Adoption Patterns

• Fastest software adoption ever seen across industries
• Legal tech (like Harvey) seeing unprecedented adoption rates in conservative sector
• Enterprise access/compliance still clunky but improving rapidly
• Context access becoming key differentiator
  • Enterprises won’t give all context to single AI tool
  • Consolidation toward job-specific AI tools with controlled context access
• Shadow IT continuing alongside official enterprise adoption

Application Layer Investment Thesis

• Many current AI startups are arbitrage plays on adoption gaps
• Non-durable unless stepping stone to deeper value
• Real opportunities exist but require honest assessment of durability
• Team quality most important factor in early evaluation
• Defensibility questions more critical than ever
• Distribution strategy essential given ease of replication

Marketplace Evolution

• Current marketplaces happening at infrastructure layer (compute, energy)
• Historical marketplaces facing uncertain future as AI changes demand patterns
• Human nature won’t fundamentally change despite AI capabilities
• People still want tactile experiences, social interactions
• Opportunity in enhancing existing behaviors vs forcing behavior change

Metrics and Evaluation Changes

• Expectations changing rapidly - million ARR no longer sufficient for some
• Focus on margin and efficiency metrics
• AI-native teams expected to achieve more with smaller engineering teams
• Milestone markers not written in stone - contextual to market moment
• Defensibility and distribution strategy scrutiny increased