Operating Models¶

March 6, 2024
in AI Strategy, Defense Technology, Operating Models, Government Technology
6 min read

The CDAO evaluation was really about decision rights

When the Pentagon announced that its inspector general was evaluating the Chief Digital and Artificial Intelligence Office in early 2024, it was tempting to ask for a verdict: Was the new organization working?

The final report produced a more revealing answer. The Department of Defense (DoD) Office of Inspector General (OIG) found that the Chief Digital and Artificial Intelligence Office (CDAO) was still operating without several foundational documents needed to make its responsibilities clear across the Department.

The problem was not a lack of ambition. CDAO had inherited data, analytics, digital services, and artificial intelligence (AI) responsibilities from four organizations. It was expected to set strategy, create policy, break adoption barriers, provide enabling services, and scale proven capabilities across one of the world's largest enterprises.

The problem was that a broad mandate does not tell thousands of people how authority should work at the boundary between organizations. Enterprise AI offices succeed or fail in those boundaries.

January 22, 2024
in Public-Sector AI, Trustworthy AI, Operating Models, AI Platforms
4 min read

Scaling Trustworthy AI in Government Requires an Operating System

Originally published in 2024; substantially revised in 2026 to deepen the analysis and incorporate additional sources.

Government agencies do not lack AI ideas. They lack repeatable mechanisms for turning a promising use case into a capability that can be evaluated, authorized, adopted, monitored, and improved.

The usual response is to scale the technology: add compute, models, data pipelines, or platform capacity. Those investments matter. But when every program defines its own risk process, evidence package, human-oversight model, security interpretation, and approval path, the organization scales experimentation while preserving the bottlenecks that prevent adoption.

Trustworthy AI scales when the enterprise standardizes the work around the model—not merely access to the model.

November 28, 2023
in Knowledge Strategy, AI Engineering, Operating Models
3 min read

Retrieval-augmented generation is not a knowledge strategy

Amazon Web Services (AWS) has made Knowledge Bases for Amazon Bedrock generally available. The service can ingest organizational documents, create a searchable vector index, retrieve relevant passages, and use them to ground a foundation model's response—with source attribution included.

Managed retrieval removes a meaningful amount of engineering work. It does not decide which organizational knowledge should be trusted.

October 26, 2023
in AI Governance, Risk Management, Operating Models
3 min read

Frontier preparedness needs decision rights before the crisis

OpenAI has announced a Preparedness team and a challenge focused on risks from increasingly capable artificial intelligence models. The effort will examine areas such as cybersecurity, persuasion, autonomy, and other severe harms, with the aim of connecting evaluation to development and deployment decisions.

Preparedness is not only the ability to detect a dangerous capability. It is the ability to decide and act while the evidence is incomplete and the stakes are rising.

October 5, 2023
in AI Strategy, Platform Engineering, Operating Models
3 min read

Inference is where AI strategy meets the budget

International Business Machines (IBM) Research has published a timely explanation of artificial intelligence inference—the moment when a trained model receives live input and produces a result. Training attracts attention because it creates the model. Inference is where the model becomes a recurring service, and where much of its lifetime cost and user experience accumulate.

For enterprise leaders, inference is not only an infrastructure concern. It is where an artificial intelligence (AI) portfolio meets a budget.

August 28, 2023
in Defense Technology, Innovation Management, Operating Models
3 min read

Replicator is an organizational test of speed and scale

The Department of Defense (DoD) has announced the Replicator initiative, an effort to field attritable autonomous systems in multiple domains at a scale of thousands within 18 to 24 months. The goal is intentionally aggressive. It is meant not only to deliver systems, but to demonstrate a repeatable way to move relevant technology into warfighters' hands faster.

Replicator will be discussed as an autonomy and manufacturing challenge. It is equally an organizational-design challenge.

May 12, 2023
in AI Strategy, Platform Engineering, Operating Models
3 min read

The enterprise AI platform is really a coordination platform

International Business Machines (IBM) has introduced watsonx, an enterprise platform that brings together a studio for foundation models, a data layer, and governance capabilities. The announcement reflects the direction many large organizations are moving: away from isolated model experiments and toward a common environment for building, adapting, and operating artificial intelligence.

The technology matters. The larger challenge is coordination.

May 1, 2023
in AI Governance, Operating Models, Organizational Learning
3 min read

Responsible AI needs distributed ownership

Microsoft's chief responsible artificial intelligence officer has published a reflection on the company's program, emphasizing leadership commitment, inclusive governance, and actionable standards. The timing is useful. Generative systems are moving into products rapidly, and many organizations are discovering that an ethics statement does not tell a product team what to do on Tuesday afternoon.

Responsible artificial intelligence (AI) needs a central function. It cannot remain the central function's job alone.