CoreWeave says the model loop in its Forge platform should be open to other clouds
CoreWeave has announced Forge, a platform linking model training, inference and evaluation so production results improve models and agents. Buyers comparing AI clouds now have a tooling claim to test next to chip capacity.
Reality
- Evidence30
- Adoption15
- Hype gap+40
- Incentives85
- Confidence40
LlamaIndex builds document models on rented GPUs that have to absorb its customers' paperwork spikes
LlamaIndex runs a workload that is 75% inference on rented GPUs, parsing millions of document pages a day that arrive in bursts. Its CEO's account at the Fully Connected event describes a buyer for whom guaranteed capacity matters more than owning hardware.
Reality
- Evidence35
- Adoption30
- Hype gap+30
- Incentives80
- Confidence40
MOUs with six large asset managers aim to mobilize more than $500 billion for AI factories. Nothing is funded yet, but the underwriting model already changes who sets compute prices.
Reality
- Evidence55
- Adoption12
- Hype gap+40
- Incentives70
- Confidence50
Oracle moves the entitlement check to the moment the query hits the database
Vipin Samar and David Knox showed an indirect prompt pulling salary data past an application's guardrails, then failing once the rule lived in the database, and argued that no agent should be trusted to police its own access.
Reality
- Evidence32
- Adoption
- Not scored
- Hype gap+35
- Incentives85
- Confidence48
Oracle pushes an agent's read limits down to the row, column and cell
Oracle's database chief says AI is turning up hundreds of security issues in code his teams spent decades hardening. His prescription for agents is authorization enforced inside the database, keyed to the end user's identity.
Reality
- Evidence38
- Adoption12
- Hype gap+30
- Incentives88
- Confidence45
Salesforce's Headless 360 opens its software to agents with no user interface
At Dreamforce 2026 Salesforce showed a way into its software with no user interface at all. The analysts who covered the event say the pricing is unsettled and the agent stack underneath is still early.
Reality
- Evidence32
- Adoption15
- Hype gap+35
- Incentives72
- Confidence42
A 27-second intrusion headlines a CrowdStrike report that averages 29 minutes
CrowdStrike's 2026 Global Threat Report carries both numbers, and the gap between them decides whether a control has to run without a person or can wait for someone on shift to look.
Reality
- Evidence46
- Adoption18
- Hype gap+42
- Incentives82
- Confidence52
CrowdStrike's SafeMind pairs attack-emulation and defensive AI models under one harness
CrowdStrike's SafeMind pairs an offensive model, Red Tempest, with a defensive one, Blue Solano, and puts both inside Falcon. The budget question that follows is who authorises the attack side and reviews what it does.
Reality
- Evidence32
- Adoption16
- Hype gap+54
- Incentives84
- Confidence46
CrowdStrike cleared its own ARR guide by 17% while revenue beat by 2%
CrowdStrike and Okta both credit enterprise agent rollouts for larger commitments. The commitments moved sharply; the quarters they reported barely did.
Reality
- Evidence58
- Adoption52
- Hype gap+28
- Incentives82
- Confidence60
The shortage is the schedule: AI capacity relief is not a 2027 line item
Dave Vellante argues HBM, advanced packaging, network fabric, power and site readiness stay tight until at least 2028. Price discovery gets pushed out with them.
Reality
- Evidence28
- Adoption37
- Hype gap+26
- Incentives71
- Confidence33