Summary
May Habib argues that enterprise AI adoption now depends less on chasing frontier models and more on fit-for-purpose harnesses, enterprise-controlled context, and cost-efficient deployment. She says buyers want context sovereignty, token costs flattened, and organizational redesign so agents can act with fewer human gates.
Key Topics
5 key topics from May Habib and Sharon Goldman at HumanX Amsterdam 2026. Thicker branches were mentioned more often in the talk.
Why Most AI Never Makes It Inside
Click a topic to open it
Key Takeaways
- Prioritize fit-for-purpose harnesses over constant frontier-model switching.
- Encode compliance, brand, and regulatory context once for reuse everywhere.
- Store enterprise context in your own cloud to preserve sovereignty.
- Redesign approval flows so agents can operate with greater autonomy.
- Evaluate open or tuned models for token cost and IP exposure.
“we can take raw intelligence and turn it into company intelligence.”
Share
Up Next
LPA
WatchNext in agenda
Where Capital & Emerging AI Categories Collide
Larry Li, Philip Chopin, Alex Hern, Shamillah Bankiya
venture capitalai startups
AKH
WatchAlso on ai agents
Past the Demo: Deploying AI Agents at Enterprise Scale
Ashwin Sreenivas, Kartik Sheth, Hugh Langley
ai agentsenterprise ai
DJL
WatchAlso on ai agents
Rewriting the Operating Model for the AI Era
Des Traynor, Jennifer Smith, Luisa Bomke
ai agentsenterprise ai