Home → Q10 Research Observatory

Q10 Research Observatory

Research policy: separate external evidence from Q10 interpretation. Dates, limitations, counterevidence and direct source links remain visible. Provenance is not the same thing as truth.

RESEARCH BRIEF 01

Adoption is broad; enterprise value is uneven.

McKinsey's 2026 survey reports nearly nine in ten respondents using AI regularly in at least one business function, while 37% report positive enterprise-level EBIT impact and only about 6% meet its high-performer definition. Q10 implication: do not sell activity as ROI; baseline the workflow and measure accepted outcomes.

~9/10regular AI useMcKinsey State of AI 2026.

37%positive EBITMcKinsey State of AI 2026.

~6%high performersMcKinsey's high-performer definition.

32%declined softwareRespondents reporting at least one purchase declined because agentic coding could build it internally.

RESEARCH BRIEF 02

Agent systems move risk from text generation into actions.

OWASP's 2026 agentic framework focuses on risks unique to autonomous applications. Security vendors are adding agent discovery, identity, permission, red-team and runtime controls. Q10 implication: basic agent security is a competitive field; Q should add vendor-neutral authority, evidence, economics, recovery and interoperability rather than pretend a kill switch alone is a moat.

RESEARCH BRIEF 03

Syn-to-syn interoperability is becoming infrastructure.

The July 2026 MCP release reported close to 500M monthly downloads across Tier-1 SDKs. A2A v1.0 provides an open language for independent agents to discover capabilities and coordinate tasks. Q10 implication: support these transports and enforce trust boundaries around them—identity, delegation, TTL, budget, evidence, termination and human escalation.

RESEARCH BRIEF 04

Provenance is necessary but does not prove a claim is true.

C2PA 2.4 supports cryptographically verifiable content provenance, machine-readable AI disclosure and additional assertions. Q10 implication: interoperate with provenance standards while maintaining a separate claim-evidence truth layer and explicit uncertainty.

RESEARCH BRIEF 05

Cost, latency and reliability are architecture variables.

AI gateways and observability platforms already provide multi-provider routing, logging, evaluation and spend controls. Q10 implication: the differentiator is not another generic dashboard; it is minimum-sufficient-intelligence routing tied to quality/safety floors, evidence and rollback.

SOURCE REGISTER

Direct sources used by the Core 50

S1McKinsey — State of AI 2026
Adoption, agents, productivity, EBIT, cost and workforce findings.S2Stanford HAI — AI Index 2026
Economy and organizational adoption evidence.S3Deloitte — Q2 2026 CFO Signals
AI cost transparency, governance and risk-management concerns.S4NIST AI RMF / AIRC
Govern · Map · Measure · Manage and TEVV resources.S5ISO/IEC 42001
AI management-system framework.S6OWASP Agentic Top 10 2026
Threats unique to agentic applications.S7Model Context Protocol
2026 specification release and ecosystem scale.S8A2A Protocol v1.0
Agent capability discovery and task collaboration.S9C2PA 2.4
Content Credentials, AI disclosure and cryptographic provenance.S10Gartner — AI spending 2026
Worldwide AI spending forecast.S11Gartner — AI platforms/models 2026
Market and buyer scrutiny of cost, latency, performance and reliability.S12Cloudflare AI Gateway
Routing, caching, observability and cost controls.S14IBM watsonx.governance
Enterprise governance and factsheet capabilities.S16Palo Alto Prisma AIRS
Agent discovery, identity, permission and runtime security.S17Arize Phoenix
AI evaluation and tracing patterns.S18OpenTelemetry
Reusable semantic conventions for telemetry.