Freddie Seba headshot

AI Keynote Speaker & Ethics Speaker & Executive Workshops

Dr. Freddie Seba

Issue #48 — Partnerships, Sovereignty, and “Agentic Coworkers”

GenAI Ethics & Governance for Leaders (Higher Ed • Healthcare • Public Interest)

Building the age of thinking machines—without outsourcing responsibility

Copyright © 2025 Freddie Seba. All rights reserved.

A small note at forty-eight

This week’s throughline is simple—and uncomfortable:

Whoever controls the stack controls the terms.

Sovereignty is becoming a strategy (and constraint). And “speed” is now a governance variable—because the path from copilots → agents is getting shorter.

The risk isn’t just a bad output.

The risk is mission drift: institutions scaling faster than their accountability systems.

Executive signal-

-Europe’s digital-stack

Push is real, but difficult: costs, fragmentation, procurement friction, and the “European champion trap” show up fast.

-Speed is now a governance variable

When low latency becomes the feature, the shift from copilots to agentic systems accelerates—so oversight has to become continuous, not periodic.

Board & trustee briefing

If you sit on a board (university, health system, nonprofit, enterprise), here are five questions to put on the agenda this quarter:

  1. Where are we becoming structurally dependent?
  2. (model family, cloud, agents, monitoring vendor)
  3. What do we measure in production—not just in pilots?
  4. (drift, safety, equity, workflow outcomes, near-misses)
  5. Who owns the incident playbook?
  6. (thresholds, escalation, rollback, communications)
  7. What is our “minimum viable transparency” promise to staff, learners, and patients?
  8. If we onboard agents, what changes in approval chains and auditability?
  9. Agents aren’t “just chatbots.”

Europe’s digital stack: sovereignty without new giants

Europe’s efforts to reduce overconcentration are wrestling with a complex design problem: how to curb dependence on US hyperscalers without recreating the same concentration with European replacements—while navigating real administrative and market friction.

Practical takeaway for institutions: treat sovereignty as a spectrum, not a slogan:

  • data residency
  • portability/exit options
  • auditability
  • procurement controls (terms, rights, and incident duties)

“Agentic AI coworkers” and the GenAI paradox

McKinsey’s year-in-review framing captures what many leaders feel: nearly everyone has invested in GenAI, but many struggle to show durable value—driving interest in agents that can transform workflows (not just answer questions).

Governance translation: if agents execute multi-step work “side by side” with people, you need controls designed for autonomy:

  • permission boundaries (recommend vs act)
  • logging and traceability
  • escalation paths
  • continuous evaluation in the workflow

Healthcare: the evidence gap (and why it matters)

A JAMA systematic review of 519 studies on testing and evaluation of health care applications of large language models highlights how heterogeneous evaluation is today—and how limited attention remains to key dimensions beyond accuracy.

Why this matters for leaders: benchmark success is not clinical readiness.

“Good at exams” doesn’t equal “safe in workflows.”

The Seba Framework: 12 Ps of responsible GenAI deployment

Use this as a pre-flight scan before scaling:

  1. Purpose — what outcome, for whom?
  2. People — ownership, approval rights, accountable executive
  3. Product — what exactly is deployed (model + workflow)
  4. Policy — rules, exceptions, escalation
  5. Privacy — minimization, retention, boundary control
  6. Permission — consent and acceptable-use limits
  7. Provenance — data lineage, grounding, traceability
  8. Performance — real-world metrics (not just demos)
  9. Protection — security, misuse testing, red-teaming
  10. Process — incident response + rollback
  11. Post-deployment — monitoring, drift, revalidation
  12. Proof — audit-ready evidence, not vibes

Closing reflection

If we’re onboarding “agentic coworkers,” we’re not just buying software.

We’re rewriting workflows, approval chains, and accountability.

Speed is impressive.

But governance is what keeps speed from turning into regret.

Gratitude (this week)

With appreciation to:

  • @University of Illinois Chicago (UIC)
  • @University of San Francisco (USF)
  • @AMIA Informatics (AMIA)
  • @Stanford HAI, @Coalition for Health AI (CHAI), @AAC&U, @AAAI—and the clinicians, students, trustees, and executives who keep sharing real-world cases and complex questions.

About the author

Freddie Seba is a lifelong learner, strategist, and academic–practitioner focused on Generative AI ethics and governance for institutional leaders. He brings 20+ years across Silicon Valley startups, corporate strategy, and graduate teaching in digital health, innovation, and GenAI ethics to help boards, executives, and faculty adopt AI responsibly and effectively.

MBA (Yale) | MA (Stanford) | Completing EdD (USF)

Speaking / Briefings: Connect on LinkedIn or visit freddieseba.com.

Transparency & Disclaimer

This newsletter is for educational and informational purposes only. It does not provide medical, healthcare, educational, instructional, accreditation, financial, investment, or professional advice. It does not create a clinician–patient, advisor–client, or instructor–student relationship. Leaders and organizations should consult appropriate professionals and institutional governance bodies before making decisions about AI deployment.

Source links are shared for credit and context; copyright belongs to the original publishers. Drafted and refined with assistive tools (including ChatGPT/GPT, Gemini, Speechify, Grammarly); synthesis, structure, and final voice remain the author’s.

References

Copyright © 2025 Freddie Seba. All rights reserved.