Antoniou Warns Shadow AI Exposes C-Suite to SOX Compliance Risks

Panel 5 - 1: From AI Adoption to AI Readiness: A Framework for Governance, Ethics, and Workforce — Ai4

Ai4

Unchecked bottom-up AI adoption exposes corporate leadership to severe regulatory liabilities because enterprises cannot retroactively govern tools they failed to prepare for.

AI-assisted summary of Panel 5 - 1: From AI Adoption to AI Readiness: A Framework for Governance, Ethics, and Workforce — original source. Check the original for full context.

Source published:

Detailed notes

## Background Published by Ai4 on YouTube on August 14, 2026, this session features George Antoniou, Full Professor of Cybersecurity at Lynn University and a Fulbright Scholar. Drawing on his field experience developing AI initiatives across academia and industry, Antoniou addresses the widening gap between rapid AI adoption and institutional readiness, introducing a practical framework for governance, ethics, and workforce development.

## TL;DR George Antoniou argues that while organizations are rapidly deploying AI tools through decentralized, bottom-up adoption, most lack the formal governance structures, policy frameworks, and executive readiness required to mitigate regulatory and operational risks. Drawing from field research, he presents a five-dimension AI Readiness Framework encompassing governance authority, ethical transparency, continuous leadership education, modernizing standard operating procedures, and workforce reskilling. Antoniou reframes workforce displacement by emphasizing that AI will shift roles toward AI-literate personnel rather than eliminate jobs outright, necessitating human-in-the-loop controls. Ultimately, he posits that while AI adoption is merely a technology decision, true AI readiness is a leadership mandate requiring standing institutional commitment.

## Atmosphere & style The presentation carries an authoritative, pragmatic, and urgent tone that blends scholarly field research with high-stakes executive consulting. Antoniou engages the room with direct, sobering challenges to leadership complacency, using vivid illustrations like executives facing compliance audits under oath and academic integrity boards paralyzed by policy vacuums. The talk feels like an incisive diagnostic briefing, shifting from sharp warnings about shadow AI risks to a structured, constructive roadmap for institutional transformation.

## Key takeaways - [00:04](https://www.youtube.com/watch?v=qivYz1lEuJE&t=4s) Organizations are rapidly adopting AI tools, yet most lack the formal readiness evaluations, governance guardrails, and authorization structures required to manage them. - [02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) AI typically enters enterprises through informal, bottom-up 'shadow AI' adoption, where unvetted tools proliferate across departments without centralized oversight or security approval. - [02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) Claiming alignment with standards like the NIST AI Risk Management Framework on paper fails without operational infrastructure and a standing AI Governance Committee with enforcement authority. - [05:47](https://www.youtube.com/watch?v=qivYz1lEuJE&t=347s) Unmonitored AI usage poses significant regulatory and compliance risks under frameworks like SOX, where C-suite executives are personally and legally accountable for data integrity. - [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) The AI Readiness Framework establishes five interconnected dimensions: AI Governance, Ethics & Transparency, Leadership Preparedness, Curriculum & SOP Alignment, and Workforce Readiness. - [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) Relying on legacy frameworks, such as outdated plagiarism or conduct policies, leaves institutions procedurally paralyzed when addressing generative AI disputes. - [11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) Ethical implementation depends on maintaining 'human-in-the-loop' and 'human-in-control' safeguards across high-stakes decision-making environments. - [11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) Because AI models and risk profiles evolve constantly, executive leadership readiness requires ongoing education rather than one-off seminars. - [11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) Rather than outright job loss, workforce displacement will affect those who do not know how to use AI, making institutional reskilling and SOP modernization essential. - [15:43](https://www.youtube.com/watch?v=qivYz1lEuJE&t=943s) Addressing fundamental governance questions unlocks tangible business outcomes, including optimized decision-making, reduced compliance exposure, and institutional trust. - [18:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1111s) While AI adoption is simply a technology decision, AI readiness is a leadership decision and responsible transformation requires an ongoing organizational commitment.

## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=qivYz1lEuJE&t=4s) Introduction and the Gap Between Adoption and Readiness

George Antoniou, Full Professor of Cybersecurity Management at Lynn University, a Fulbright Scholar, and a Fulbright Ambassador, introduces his session on bridging the gap between organizational AI adoption and AI readiness. The framework presented is not an abstract academic exercise or a mere literature review; it is the culmination of a five-month field residency conducted at the University of Tirana in Albania in 2025. Across two universities and two countries, Antoniou worked to synthesize real-world implementations into a structured model for governance, ethics, leadership, and workforce capability.

```mermaid flowchart TD accTitle: The AI Readiness Framework Overview subgraph Core["Core Philosophy"] CR["AI Readiness Framework<br/>'You cannot govern what you have not prepared for'"] end

subgraph Dimensions["Five Interconnected Dimensions"] D1["AI Governance & Authority<br/>Standing Committee"] D2["Ethics & Transparency<br/>Human-in-the-Loop"] D3["Leadership Preparedness<br/>Continuous Education"] D4["Curriculum & SOP Modernization<br/>Embedded AI Workflows"] D5["Workforce Readiness<br/>Literacy & Reskilling"] end

CR --- D1 CR --- D2 CR --- D3 CR --- D4 CR --- D5 ```

While organizations across every sector are aggressively adopting, deploying, and scaling artificial intelligence, Antoniou challenges the audience on whether they are truly prepared for what follows. He notes that while almost everyone is using AI as a tool, very few institutions have paused to undergo a formal evaluation, establish compliance guardrails, or determine how decisions surrounding the technology are authorized and managed.

---

[02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) The Shadow AI Dilemma and the Organizational Vacuum

The core issue facing modern enterprises is not technological, but organizational. In almost every environment—spanning academic institutions, government agencies, and corporate enterprises—AI enters through an informal, bottom-up pathway. An individual employee or faculty member discovers an AI tool, finds it useful, and shares it with colleagues. Soon, entire departments, faculty bodies, and student populations are relying on software that no central authority ever formally vetted, budgeted for, or approved.

This unchecked adoption introduces massive vulnerabilities in Governance, Risk, and Compliance (GRC). Antoniou points out that although many organizations claim to align with frameworks like the NIST AI Risk Management Framework (AI RMF), they lack the operational infrastructure to enforce them. When asked basic questions—such as who is legally responsible for AI in the company, how many tools are currently deployed, or whether a formal AI governance committee exists—most leaders have no concrete answers. Instead of standing governance bodies with decision-making power, organizations rely on informal, toothless task forces.

- **Unvetted Proliferation:** Tools are adopted virally across teams without centralized purchasing, risk scoring, or security authorization. - **NIST AI RMF Gaps:** Organizations frequently cite NIST standards on paper while missing the foundational governance policies required to operationalize them. - **Committee vs. Task Force:** An ad-hoc task force lacks the permanent institutional mandate and authority needed to enforce compliance; a standing AI Governance Committee is required.

---

[05:47](https://www.youtube.com/watch?v=qivYz1lEuJE&t=347s) Regulatory Scrutiny, Compliance, and Executive Accountability

Antoniou emphasizes that compliance is the ultimate test of organizational readiness. For publicly traded corporations governed by Sarbanes-Oxley (SOX) and New York Stock Exchange regulations, the CEO, CFO, and executive leadership are personally and legally liable for the integrity of their data systems and financial reporting. When regulatory audits arrive, executives cannot afford to be ignorant of the AI tools operating inside their perimeter.

If C-suite executives were asked under oath to certify exactly which AI tools their employees use and whether those systems produce auditable, uncompromised outputs, the vast majority would fail. The sheer lack of centralized inventory and approval processes turns AI from a productivity booster into a serious regulatory liability. True organizational readiness means having the governance and technical controls in place before an external audit or compliance review forces the issue.

---

[08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) The 5-Dimension Framework and Establishing Governance Authority

The foundational premise of Antoniou’s model is straightforward: *"You cannot govern what you have not prepared for."* If an enterprise fails to prepare for AI adoption, attempting retroactive governance over hundreds of decentralized tools is nearly impossible. Readiness requires establishing clear lines of ownership, accountability, and authorization through five interconnected dimensions that can be implemented concurrently.

| Framework Dimension | Core Focus | Key Institutional Mechanism | |:--- |:--- |:--- | | **1. AI Governance** | Authority & Accountability | Standing AI Governance Committee with policy enforcement power | | **2. Ethics & Transparency** | Responsible Use & Disclosure | Human-in-the-loop oversight; auditable decision criteria | | **3. Leadership Preparedness** | Strategic Literacy | Continuous executive education beyond one-time workshops | | **4. Curriculum & SOP Alignment** | Workflow Modernization | Embedding AI guidelines directly into course curricula and SOPs | | **5. Workforce Readiness** | Reskilling & Adaptation | Comprehensive AI literacy programs to prevent job displacement |

To illustrate the cost of unpreparedness, Antoniou highlights a common dilemma in higher education. A professor catches a student submitting an AI-generated paper and attempts to discipline them. The student pushes back, asking for the specific institutional policy forbidding AI usage. When the faculty member consults the administration, they discover only a legacy plagiarism policy, which fails to legally or procedurally cover generative text. Academic integrity boards then become bogged down in endless debates over whether AI detection software is accurate, simply because the institution adopted technology without defining policies or authorization channels first.

---

[11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) Ethics, Leadership Education, and the Workforce Reskilling Shift

Moving through the dimensions, Antoniou addresses the ethical component, emphasizing transparency and the necessity of human oversight. Whether evaluating an academic paper or making diagnostic healthcare decisions, systems must incorporate "human-in-the-loop" and "human-in-control" mechanisms. For organizational leaders, readiness is not about monitoring compute token expenses; it is about deep, continuous executive education. AI models, versions, and risk profiles evolve daily, meaning a one-off seminar is wholly insufficient for executives tasked with making strategic risk decisions.

Addressing workforce anxieties, Antoniou reframes the displacement narrative:

> "AI is not going to take anybody's job. The jobs are going to be shifted... Who is going to be replaced by AI is people that they don't know how to use AI in five years."

He points to historical shifts in labor: Manhattan transitioned from horse-drawn carriage drivers and street sweepers to taxi drivers, and is now transitioning toward autonomous vehicles like Waymo. However, autonomous Waymo vehicles do not operate in a vacuum; remote operators sit inside control rooms ready to take manual control if a malfunction occurs. Similarly, enterprises must prioritize aggressive workforce reskilling and AI literacy so employees can operate alongside modern tools rather than be replaced by AI-literate competitors. Standard Operating Procedures (SOPs) and university curricula must be fundamentally rewritten to embed AI into everyday workflows.

---

[15:43](https://www.youtube.com/watch?v=qivYz1lEuJE&t=943s) Five Strategic Questions and the Business Impact of Readiness

To successfully operationalize the AI Readiness Framework, leadership teams must be capable of providing explicit answers to five fundamental questions:

```mermaid flowchart LR accTitle: Five Core Questions to Strategic Outcomes subgraph Inputs["5 Strategic Questions & Solutions"] Q1["Who owns AI? <br/>➔ Governance Committee"] Q2["What principles guide us? <br/>➔ Ethical Transparency"] Q3["Are leaders prepared? <br/>➔ Executive Education"] Q4["Is the workforce ready? <br/>➔ AI Literacy & Reskilling"] Q5["Are SOPs/curricula aligned? <br/>➔ Modernized Procedures"] end

subgraph Outputs["Organizational Outcomes"] R1["Optimized Decision-Making"] R2["Reduced Compliance Risk"] R3["Heightened Institutional Trust"] R4["Workforce Confidence"] R5["Responsible Innovation"] end

Inputs --> Outputs ```

When an organization successfully implements these five dimensions, Antoniou explains that it unlocks five tangible organizational outcomes: sharper decision-making, significantly reduced operational and compliance risk, stronger trust among stakeholders, increased workforce confidence, and a sustainable culture of responsible innovation.

---

[18:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1111s) Three Guiding Principles for Responsible AI Transformation

Antoniou concludes his presentation by leaving the audience with three core principles to guide their institutional strategy moving forward:

1. **"AI adoption is a technology decision."** It determines which specific tools or models an organization chooses to deploy. 2. **"AI readiness is a leadership decision."** It establishes whether the organization possesses the strategic literacy, governance structures, and risk maturity to manage that technology. 3. **"Responsible AI transformation is an organizational commitment."** It requires an ongoing cultural and operational investment to align workforce skills, ethics, and procedures with technological progress.

Ultimately, only organizational leaders themselves can answer whether their enterprise is genuinely ready for the technology it is already using. Antoniou invites attendees to connect with him on LinkedIn and visit his research poster to continue the conversation on AI readiness and governance.

## Notable quotes > [00:00](https://www.youtube.com/watch?v=qivYz1lEuJE&t=0s) If your organization is adopting AI faster than it can govern it, this session is for you. — George Antoniou

> [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) You cannot govern what you have not prepared for. — George Antoniou

> [11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) AI is not going to take anybody's job. The jobs are going to be shifted... — George Antoniou

> [11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) Who is going to be replaced by AI is people that they don't know how to use AI in five years. — George Antoniou

> [11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) human-in-the-loop — George Antoniou

> [11:46](https://www.youtube.com/watch?v=qivYz1lEuJE&t=706s) human-in-control — George Antoniou

> [18:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1111s) AI adoption is a technology decision. — George Antoniou

> [18:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1111s) AI readiness is a leadership decision. — George Antoniou

> [18:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1111s) Responsible AI transformation is an organizational commitment. — George Antoniou

## Mentioned George Antoniou, Lynn University, Fulbright Program, University of Tirana, Albania, NIST AI Risk Management Framework (AI RMF), Sarbanes-Oxley (SOX), New York Stock Exchange (NYSE), Waymo, Manhattan, LinkedIn, Ai4

## Video description George Antoniou, Full Professor Cybersecurity, Lynn University

Organizations are adopting AI fast. But are they ready to govern it responsibly? This session examines the gap in AI adoption and organizational readiness, drawing on my Fulbright Scholar experience developing AI and cybersecurity initiatives with academia, government, and industry. I will introduce the AI Readiness Framework, a practical model for governance, ethics, leadership, and workforce readiness. You will leave with practical strategies to strengthen AI governance and readiness. If your organization is adopting AI faster than it can govern it, this session is for you.

[Source](https://www.youtube.com/watch?v=qivYz1lEuJE)

中文简报

安东尼奥警告：影子AI使企业高管面临SOX合规风险 缺乏约束的自下而上式AI采用令企业领导层面临严峻的监管责任，因为企业无法对其未提前做好准备的工具进行追溯性治理。 George Antoniou指出，尽管各类组织正在通过分散、自下而上的方式快速部署AI工具，但多数机构缺乏缓解监管与运营风险所需的正式治理结构、政策框架以及高管准备度。结合实地调研，他提出了一个涵盖治理权威、伦理透明度、持续领导层教育、标准作业程序现代化以及员工技能重塑的五维AI准备度框架。Antoniou重新界定了劳动力替代问题，强调AI会将岗位转向具备AI素养的人才，而非彻底消除就业，因此必须建立“人在回路”的控制机制。他最终指出，采用AI仅仅是一项技术决策，而真正的AI准备度则是一项领导层职责，需要组织建立长期的制度承诺。 多数机构在未建立正式治理评估与审批架构的情况下正面临“影子 AI（Shadow AI）”野蛮生长的风险，未经审查的工具在部门间自下而上泛滥，导致 NIST 等风险管理框架流于形式。 (2:31) 在受 SOX 等法案监管的环境中，未受监控的 AI 带来巨大的合规隐患，若高管无法确凿证明内部所用 AI 工具的输出可审计且未受篡改，将面临直接的个人法律责任。 (5:47) Antoniou 提出了涵盖 AI 治理、伦理透明度、领导力就绪、SOP 与课程对接以及劳动力重塑的五维“AI 就绪度框架”，强调依赖过时制度将使机构在应对生成式 AI 争议时陷入程序瘫痪。 (8:31) 企业必须确立“人在回路（Human-in-the-loop）”机制并开展持续的高管教育；AI 并不会直接消灭岗位，而是淘汰不会使用 AI 的从业者，因此通过重写 SOP 与全员技能再培训才能实现真正的组织就绪。 (11:46)

Reading summary

George Antoniou argues that while organizations are rapidly deploying artificial intelligence, true institutional capability requires proactive 'AI readiness'—establishing standing governance, executive literacy, and operational controls before adoption rather than attempting retroactive oversight. Bottom-Up Adoption and Regulatory Exposure (2:31) Antoniou observes that enterprise AI adoption rarely begins through structured procurement; instead, it proliferates through bottom-up channels as individual employees and faculty independently adopt unvetted tools. While many organizations claim alignment with external standards such as the NIST AI Risk Management Framework, they frequently lack operational infrastructure, centralized tool inventories, and standing governance bodies, relying instead on ad-hoc task forces that lack enforcement power. This unmanaged proliferation creates severe compliance vulnerabilities, particularly for publicly traded entities governed by Sarbanes-Oxley (SOX) and New York Stock Exchange regulations. Under these frameworks, corporate officers face personal and legal liability for system integrity and financial reporting. Antoniou asserts that most C-suite executives would currently fail if asked under oath to identify all AI tools active within their enterprise perimeter and certify their outputs as auditable, transforming shadow AI from a productivity aid into a direct regulatory liability. Institutional Readiness and the Governance Framework (8:31) Drawing from a five-month field residency in 2025 across two universities in Albania (including the University of Tirana) and the United States, Antoniou formulated a five-dimension readiness framework based on the premise that organizations cannot govern what they have not prepared for. The model prescribes five concurrent institutional mechanisms: a standing AI governance committee with policy enforcement authority, transparent ethical safeguards with human oversight, continuous executive education, modernized standard operating procedures (SOPs) and curricula, and systematic workforce reskilling. To illustrate the hazard of deploying tools before establishing rules, Antoniou cites higher education disciplinary disputes where professors sanction students for submitting AI-generated coursework. When students appeal, institutions often discover that legacy plagiarism policies fail to cover generative models legally or procedurally. As a result, academic integrity boards become paralyzed by debates over AI detection tool reliability, illustrating how premature adoption without updated governance policies creates institutional friction. Human Oversight and the Workforce Reskilling Shift (11:46) Within the ethical and operational dimensions, Antoniou stresses that critical evaluations—ranging from grading to healthcare diagnostics—must retain 'human-in-the-loop' and 'human-in-control' safeguards. Addressing labor concerns, he rejects the forecast that AI will cause widespread net unemployment, forecasting instead that labor will undergo a structural shift in which workers who fail to learn AI tools within five years will be displaced by AI-literate professionals. Antoniou contextualizes this shift through urban transport history, observing how Manhattan progressed from horse-drawn carriages to taxis, and subsequently toward autonomous vehicles like Waymo. He notes that autonomous fleets do not operate without humans; remote technicians remain stationed in control rooms to assume manual control during system failures. Consequently, organizations must rewrite SOPs and curricula to integrate AI into routine workflows, ensuring the workforce is continuously reskilled to oversee automated systems. Strategic Accountability and Core Guiding Principles (15:43) To transition from passive adoption to readiness, Antoniou states that executive teams must resolve five foundational questions: defining explicit ownership of AI, identifying governing ethical principles, verifying leadership and workforce competencies, and aligning operational procedures. Resolving these operational questions enables institutions to mitigate compliance risks, strengthen stakeholder trust, and improve organizational decision-making. Antoniou concludes by differentiating three governing concepts: AI adoption is merely a technical decision regarding specific software deployments; AI readiness is a leadership decision concerning governance infrastructure and risk literacy; and responsible transformation is an organizational commitment requiring sustained cultural and financial investment across workflows. Antoniou concludes that whether an enterprise is genuinely prepared to govern the artificial intelligence systems already operating within its perimeter remains an internal leadership question, emphasizing that permanent governance structures and workforce training must precede further technological expansion.

精炼阅读摘要

林恩大学网络安全管理学正教授乔治·安东尼奥（George Antoniou）基于其2025年跨国驻留研究指出，组织技术采纳与实际就绪度之间存在显著断层；AI采纳本质上是技术工具的选择，而AI就绪度则是一项领导力与组织治理决策，企业必须在采纳前建立具备强制力的治理体系与流程规范。 影子AI蔓延与治理真空 (2:31) 在高校、政府机构与现代企业中，人工智能主要通过自下而上的非正式路径渗透。个别员工或教师自行试用外部工具并进行同业分享，最终演化为部门甚至全校范围对未经审查软件的依赖。此类工具既未经集中化预算审批，也未经技术风险评估与安全授权。这种“影子AI”的自发式扩散，导致组织在治理、风险与合规（GRC）层面暴露出系统性脆弱性。 这种治理真空进一步体现在标准落实的虚化上。尽管不少机构声称遵循美国国家标准与技术研究院人工智能风险管理框架（NIST AI RMF），但在实际操作中缺乏落地执行的组织基础设施。当面临界定AI法律责任主体、清点已部署工具数量或确认决策授权等基本问询时，多数领导层无法给出具体答案。许多组织倾向于设立临时性工作组，但安东尼奥强调，缺乏正式制度授权的临时架构无法强制推行合规要求，机构必须设立具备实质决断权的常设AI治理委员会。 高管合规责任与审计风险 (5:47) 合规性是检验组织AI就绪度的最终标准。安东尼奥强调，对于受萨班斯-奥克斯利法案（SOX）和纽约证券交易所（NYSE）监管的上市企业而言，首席执行官（CEO）、首席财务官（CFO）及高管团队对数据系统完整性与财务汇报结果承担直接的个人与法律责任。一旦遭遇外部监管审计，高管团队无法以内置AI系统运作细节不透明为由推卸责任。 安东尼奥提出评估假设：若要求企业高管在宣誓状态下明确说明员工正在使用的所有AI工具清单，并证明这些工具产出的数据未受污染且具备可审计性，绝大多数高管将无法通过验证。由于组织缺乏中心化的AI资产清单与审批流程，未经管理的AI工具极易从提高生产力的辅助手段演变成严重的合规负债。因此，真正的就绪度要求企业在面临外部审计或合规核查之前，先期建立技术控制措施与治理防线。 五维就绪度框架与规程重构 (8:31) 安东尼奥框架的核心哲学是“你无法治理未曾做好准备的事物”。如果企业在采纳前缺乏前瞻准备，事后对数百种分散工具进行追溯性治理几乎无法实现。为此，他在两所大学与跨国实践中总结出涵盖五个维度的协同框架：确立具备强制执行力的常设AI治理委员会（AI治理）；保障人机协同与决策透明度（伦理与透明）；推行管理层持续教育而非一次性研讨会（领导力就绪）；将AI规程深度嵌入日常业务流程（标准作业程序与课程规程更新）；实施全面的素养培训以防人员脱节（劳动力就绪）。 高等教育领域的学术诚信争议典型地反映了制度脱节的代价。当教师发现学生提交AI生成的论文并试图依规处分时，学生反向索要禁止使用AI的具体校规。校方介入后往往发现现存规程仅适用于传统抄袭，在法律与程序上均未界定生成式文本的属性。结果，学术诚信委员会陷入对AI检测软件准确率的旷日持久争论之中。这一现象证明，在未设立明晰政策与授权通道前盲目采纳新技术，必然导致机构管理陷入被动追责的困境。 人机在环机制与劳动力技能重塑 (11:46) 在伦理与执行层面，安东尼奥主张无论是评估学术成果还是提供医疗诊断决策，系统必须强制实施“人在环路”（human-in-the-loop）或“人在控制”（human-in-control）机制。对于管理层而言，AI就绪度不能局限于监控计算令牌（tokens）的开销，而必须进行持续性的高管认知迭代，以应对日常更新的模型版本与风险特征。 针对劳动力可能被技术取代的担忧，安东尼奥提出重构性判断：AI本身不会直接消灭工作，但在未来五年内，不掌握AI工具的人将被熟练使用AI的劳动者所取代。他以曼哈顿交通史为例，指出交通工具从马车向出租车乃至Waymo自动驾驶演进的过程中，自动驾驶车辆并非在真空中完全自治，而是由控制室内的远程操作员随时准备人工接管。企业必须推动员工技能重塑与AI素养培训，同时彻底重写标准作业程序（SOP）和教学大纲，使劳动力能够在人机协作环境中有效履职。 安东尼奥总结道，AI采纳是关于工具选择的技术决策，而AI就绪度与组织转型则是关乎战略素养、制度流程与治理成熟度的领导力承诺。然而，面对去中心化的实际使用现状与快速迭代的外部工具，组织究竟能否通过明确问责主体与重构业务流程来真正化解治理真空，仍取决于管理层是否具备长期投入制度资本的决心。

Detailed notes — zh

## Background 该视频由 Ai4 于 2026 年 8 月 14 日发布在 YouTube 平台。内容由林恩大学（Lynn University）网络安全管理正教授兼富布赖特学者 George Antoniou 主讲，结合其实地驻留研究经验，重点剖析了组织快速采用 AI 与治理就绪度脱节的问题，并提出了涵盖治理、伦理、领导力与劳动力就绪的实用框架。 ## TL;DR 演讲指出，当前组织普遍存在自下而上无序扩散使用 AI 但缺乏审批授权与问责机制的治理盲区。为了解决“无法治理未做好准备的事物”这一困境，Antoniou 教授提出了包含常设治理委员会、伦理与人在回路机制、领导力持续教育、SOP现代化以及劳动力技能重塑等五个维度的“AI 就绪度框架”。他强调 AI 岗位正在经历结构性转移而非简单淘汰劳动力，并总结出“AI 采纳是技术决策，AI 就绪是领导力决策，而负责任的 AI 转型是一项组织承诺”。 ## Atmosphere & style 演讲整体呈现出严谨务实、条理分明且极具针对性的专业基调。主讲人依托学术实战与国际项目调研经验，语言干练笃定，通过对比传统企业法案合规以及曼哈顿交通演变等生动案例，层层剥离技术狂热背后的管理真空。现场氛围既带有对合规风险与问责缺失的尖锐警示，又充满了为决策层提供清晰落地行动纲领的建设性力量。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=qivYz1lEuJE&t=4s) AI 就绪度框架源自富布赖特学者在阿尔巴尼亚为期五个月的驻留实战经验，强调跨国、跨机构的实操落地而非脱离实际的纯理论文献综述。 - [02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) 当前组织普遍存在 AI 采用速度飞快与治理就绪度严重滞后的鸿沟，员工自下而上自发使用工具，却缺乏正规审批与合规评估流程。 - [02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) 尽管许多企业声称遵循 NIST AI RMF 等合规框架，但缺乏明确的前置就绪机制和法定责任人，高管无法像对财务报表那样对内部 AI 工具合规性签字作保。 - [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) AI 就绪度框架的核心哲学是“你无法治理你未曾做好准备的事物”，就绪度的本质是确立明确的问责机制。 - [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) 组织必须成立拥有决策实权的常设 AI 治理委员会，而不是形同虚设的临时特别工作组（Task Force），以确保责任落实到人。 - [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) 在高风险决策场景（如医疗诊断）中，必须建立伦理与透明度机制，引入“人在回路”（Human-in-the-loop）确保人类专业把关权。 - [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) 领导力就绪要求管理者超越对 Token 成本等财务数字的单一关注，通过系统化、持续性的高管教育深入理解 AI 业务定位与合规设计。 - [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) 无论是学术界的课程大纲还是企业界的标准作业程序（SOP），都必须顺应 AI 时代完成现代化重构并与流程规范深度融合。 - [14:10](https://www.youtube.com/watch?v=qivYz1lEuJE&t=850s) AI 不会直接淘汰人类，但不会使用 AI 的人会被熟练使用 AI 的人取代；劳动力正像曼哈顿交通从马车演进到无人驾驶那样，转向高阶监控与协作岗位。 - [15:43](https://www.youtube.com/watch?v=qivYz1lEuJE&t=943s) 构建 AI 就绪框架必须明确回答五个核心问题：谁掌管 AI、遵循什么伦理准则、领导层是否持续学习、劳动力是否具备 AI 素养并重塑技能、SOP/课程是否完成对齐。 - [18:27](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1107s) AI 采纳（Adoption）是一项技术决策，AI 就绪（Readiness）是一项领导力决策，而负责任的 AI 转型（Responsible Transformation）则是一项组织承诺。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=qivYz1lEuJE&t=4s) 引言与研究背景：从阿尔巴尼亚富布赖特项目谈起 演讲者 George Antoniou 教授向与会者致意并介绍自己的背景。他是林恩大学（Lynn University）网络安全管理专业的正教授、富布赖特学者（Fulbright Scholar）及近期获任的富布赖特大使。该研究成果源于他在 2025 年前往阿尔巴尼亚地拉那大学（University of Tirana）进行为期五个月的富布赖特学者驻留研究项目。他特别强调，即将展示的框架并非脱离现实的纯文献综述（Literature Review），而是历时五个月、跨越两个国家和两所大学的实地实操与落地验证结晶。 当下全球都在快速推进人工智能的应用，但组织在拥抱技术的同时，往往忽略了最核心的治理底线。大家都在使用、部署甚至规模化扩展 AI，然而几乎没有组织真正评估过自身是否在合规、风险与管理层面做好了准备，更鲜有组织通过正规严谨的决策流程来确定如何使用这项工具。Antoniou 指出，本次演讲的核心议题正是“AI 采用的超快速度”与“组织就绪度（Readiness）滞后”之间的巨大鸿沟——我们究竟有没有为 AI 做好准备？ - **研究背景**：基于 2025 年阿尔巴尼亚地拉那大学的富布赖特 5 个月驻留资助项目。 - **跨国协作**：由 2 所大学跨越 2 个国家联合开展实地研究，侧重实战落地而非单纯理论推导。 - **核心追问**：企业与院校在规模化应用 AI 之前，是否经历过正式的评估与决策治理流程？ ```mermaid flowchart TD accTitle: AI Adoption vs AI Readiness Dilemma accDescr: 展现从自发采用到合规与治理滞后的现实差距 subgraph S1["现实采用现状 (Adoption)"] A["员工 / 师生自发引入工具"] --> B["自下而上快速扩散传播"] B --> C["缺乏正式授权与预算审批"] end subgraph S2["治理与合规缺口 (Governance Gap)"] D["未定义问责人与政策"] E["GRC / 合规审计风险暴露"] F["缺少常设 AI 治理委员会"] end S1 -.->|"速度严重脱节"| S2 ``` --- ### [02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) AI 采用与组织就绪度之间的脱节与合规风险 AI 的发展极其迅猛，各家组织都在试图迎头赶上，有的在快步走，有的甚至还在艰难爬行。然而，AI 带来的真正挑战并非单纯的技术故障，而是深刻的组织管理问题。在大学、政府机构以及各类企事业单位中，普遍存在一种自下而上的无序扩散模式：某个人偶然发现了一款好用的 AI 工具并开始使用，随后分享给同事，接着教职员工开始使用，学生和普通员工也全面跟进。在这一连串的链式传播中，最关键的问题却被遗漏了——“究竟是谁批准了这项工具？谁给出的授权？它走过了什么审批流程？” 在治理、风险与合规（GRC）受到严格监管的行业中，这种无序状态尤为危险。当合规审计或外部评估到来时，管理层往往才会惊恐地意识到组织根本没有考虑过治理问题。虽然美国国家标准与技术研究院推出了专门的风险管理框架（NIST AI RMF），许多企业也声称自己在遵循该标准，但他们普遍缺失了关键的一环——前置的就绪度。当被问及“谁在对组织内的 AI 负责”时，几乎没有哪家机构能立刻给出一个明确的名字，也没有设立具备实质权力的常设治理委员会，充其量只有临时的特别工作组（Task Force）。 与传统企业合规标准相比，这种混乱形成了鲜明反差。在《萨班斯-奥克斯利法案》（Sarbanes-Oxley）或纽约证券交易所的监管要求下，上市公司首席执行官（CEO）和首席财务官（CFO）必须对财务报表与数据完整性承担法定背书责任；但若询问这些高管，是否敢为公司内部员工使用的每款 AI 工具的合规性签字作保，多数人根本无法回答，因为企业内部充斥着大量未获批准、未经备案的影子 AI 工具。 - **自发采用的无序性**：缺乏统一审批流程，工具在组织内部呈“野蛮生长”态势。 - **框架落地缺失**：尽管参考了 NIST AI RMF，但多数组织缺乏前置就绪机制与明确的责任人。 - **合规问责断层**：高管无法像应对 Sarbanes-Oxley 财报合规那样，确保企业内部 AI 工具使用合法合规。 --- ### [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) AI 就绪度框架与核心维度：治理、伦理、领导力与课程现代化 为了弥补这一鸿沟，Antoniou 提出了包含五个相互关联维度的“AI 就绪度框架”（AI Readiness Framework）。该框架的核心哲学十分简单直接："You cannot govern what you have not prepared for."（你无法治理你未曾做好准备的事物）。就绪度的本质就是建立明确的问责制（Accountability）。组织必须成立具备决策与管理实权的常设 AI 治理委员会，而不是形同虚设的临时专项小组（Task Force），以明确究竟由谁来对 AI 决策及其引发的潜在事故承担最终责任。 ```mermaid flowchart TD accTitle: AI Readiness Framework Five Dimensions accDescr: 展示 AI 就绪度框架的五个互联维度 Center["AI 就绪度框架<br/>(AI Readiness Framework)"] Center --- D1["维度 1：治理与常设委员会<br/>(Governance & Accountability)"] Center --- D2["维度 2：伦理与透明度<br/>(Ethics & Human-in-the-Loop)"] Center --- D3["维度 3：领导力准备度<br/>(Leadership Preparedness)"] Center --- D4["维度 4：课程与 SOP 现代化<br/>(Curriculum / SOP Modernization)"] Center --- D5["维度 5：劳动力期望契合<br/>(Workforce Alignment)"] ``` 缺乏治理政策与就绪度会导致具体的管理冲突。以学术界为例，当教师发现学生使用 AI 撰写论文并提出指控时，学生会反问“学校哪条政策规定不能用 AI 写论文？”；教师向管理层求助时，才发现学校只有针对传统学术抄袭（Plagiarism）的条例，并没有专门的 AI 使用政策，最终陷入 AI 检测工具是否准确、学生是否违规的无休止争论中。同样的窘境每天也在企业员工使用 AI 处理日常业务中上演。 围绕该框架，Antoniou 逐一拆解了几个关键维度的内涵： 1. **治理与权威（Governance & Authority）**：建立具有实质授权的常设 AI 治理委员会，明确责任人和审批制度，确保每款工具的使用都责任到人。 2. **伦理与透明度（Ethics & Transparency）**：界定使用 AI 的伦理边界与透明机制。在医疗等高风险场景中，AI 无法直接做出最终决策，必须引入“人在回路”（Human-in-the-loop）机制来确保专业人士的实质把关。 3. **领导力就绪度（Leadership Preparedness）**：管理者若想推动 AI 落地，不能只停留在口头意愿或关注 Token 花费等财务数字上，而必须通过系统化的持续培训（超越简单的单次研讨会），真正理解 AI 在业务流中的定位与合规政策设计。 4. **课程与标准作业程序现代化（Curriculum & SOP Modernization）**：学术界的课程大纲需要顺应 AI 时代完成现代化重构；对应到企业界，各类标准作业程序（SOP）也必须将 AI 深度融入流程规范，并与劳动力市场的真实技能期望保持紧密契合。 - **核心箴言**：“你无法治理你未曾做好准备的事物”（"You cannot govern what you have not prepared for."）。 - **常设委员会 vs 临时小组**：必须设立拥有实质权力的常设治理委员会（Standing Committee），而非权责不明的 Task Force。 - **伦理与人在回路**：高风险决策（如医疗诊断）必须保留人类最终把关权（Human-in-the-loop）。 - **制度现代化**：推动高校课程大纲与企业 SOP 针对 AI 时代进行全面重构与对齐。 ### [14:10](https://www.youtube.com/watch?v=qivYz1lEuJE&t=850s) 劳动力就绪与岗位变迁：从马车到自动驾驶的人在回路 组织在推行人工智能时，必须明确员工、教职员工、学生以及客户的真实诉求。AI 就绪框架中的核心要素之一便是“劳动力就绪”（Workforce Readiness）。许多人常常陷入恐慌，担心自己会被 AI 彻底淘汰，但现实并非如此简单。真正会在未来五年内被取代的，是那些“不会使用 AI 的人”。AI 本身不会夺走任何人的工作，工作岗位的本质是在经历结构性转移（Shifting）。 回顾历史，以纽约曼哈顿的交通演变为例：早期街道上充斥着马车和清扫路面的工人；随后时代变迁，出租车和专职司机取代了马车；而如今，Waymo 等自动驾驶汽车逐渐普及，传统司机的身影可能不再常见，但并不意味着人类的缺席。每一辆 Waymo 无人车的背后，都有一个配备专业监控人员的控制中心。这些车辆并非完全无人管辖在街头行驶，一旦车辆发生故障或遭遇突发状况，控制人员随时准备接管控制权。这就是典型的“人在回路”（Human-in-the-loop）与人类控制机制，也是劳动力向高阶技能与协同岗位转变的生动写照。 ```mermaid flowchart LR accTitle: 劳动力技能演进模型 subgraph S1["传统模式"] A["传统劳动力 / 司机"] end subgraph S2["AI 技术驱动"] B["自动化工具 / 自动驾驶"] end subgraph S3["新形态劳动力"] C["具备 AI 素养的人才"] D["人在回路监管与协作"] end A --> B B --> C B --> D ``` --- ### [15:43](https://www.youtube.com/watch?v=qivYz1lEuJE&t=943s) 构建 AI 就绪框架必须回答的五个核心问题 为了真正实现组织就绪，管理层必须清晰回答关于 AI 治理的五个根本问题： 1. **谁来掌管 AI（Who owns AI）？** 答案是组织的“治理委员会”（Governance Committee）。唯有设立跨职能的治理委员会，才能确保 AI 在整个组织内得到规范管理与统筹权责。 2. **遵循什么指导原则（What principles guide us）？** 答案是“伦理道德”（Ethics）。伦理的本质在于扬善与做正确的事（Doing the right thing），并在技术应用中保持高度透明。 3. **领导层是否做好准备（Are your leaders prepared）？** 答案是“持续的高管教育”（Executive Education）。这绝不是派领导参加一次研讨会或工作坊就能解决的，而是一个日复一日的持续学习过程。AI 工具与底层模型每天都在快速迭代，领导层必须与员工同步接受持续培训。 4. **劳动力是否就绪（Is your workforce ready）？** 核心在于“AI 素养”（AI Literacy）与“技能重塑”（Reskilling）。AI 素养意味着员工深刻理解 AI 工具的功能与局限，而技能重塑则是组织必须提供的兜底保障。如果不进行再培训，员工就会输给那些熟练掌握 AI 工具的竞争对手。 5. **课程或标准操作程序（SOP）是否对齐？** 无论是在学术界还是工业界，都必须推进现代化改革，将 AI 深度嵌入到教学大纲或企业的标准操作规程中。 当组织能够圆满回答这五个问题并付诸实践时，将直接收获五大核心价值：做出更明智的决策、显著降低潜在风险、建立起内外部信任、打造充满自信的现代化劳动力团队，并最终构建一个能够承担创新责任的组织。 - **AI 责任归属**：由专设的 AI 治理委员会统一负责与监督 - **伦理准则**：以行善、透明度以及合规操作为核心指引 - **高管教育机制**：打破“一次性培训”思维，转向常态化、持续性的高阶知识更新 - **员工赋能路径**：通过普及 AI 素养与专项 Reskilling，避免员工因技能断层而被淘汰 - **流程与大纲重构**：将 AI 应用规范与能力要求全面写入组织 SOP 与学术课程体系 --- ### [18:27](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1107s) AI 采纳、就绪与责任转型的三大定论 许多组织虽然已经走在采纳 AI 的快车道上，但普遍缺乏支撑其稳健运转的“就绪框架”。关于 AI 的组织落地，有三条不可混淆的根本论断： 首先，**AI 采纳（AI Adoption）是一项技术决策**。每个组织在当下或者不久的将来，都需要决定采购和部署哪一款或哪几款具体的 AI 工具。 其次，**AI 就绪（AI Readiness）是一项领导力决策**。组织是否真正做好就绪准备，完全取决于领导层的战略眼光、制度建设与前瞻指引，绝非普通员工单方面能够决定的事项。 最后，**负责任的 AI 转型（Responsible AI Transformation）是一项组织承诺**。唯有组织从顶层设计上坚定拥抱 AI 治理、伦理与全员重塑，上述所有关键要素才能真正落地生根。 > "AI adoption is a technology decision... AI readiness is a leadership decision... and responsible AI transformation is an organizational commitment." 演讲最后，George Antoniou 教授向全场提问：“你们准备好了吗？”他指出这一问题的答案只能由各组织自行解答，若尚未准备妥当，则应尽快引入该 AI 就绪框架。随后，他感谢了现场观众，并邀请对该议题感兴趣的参会者在会后前往展位交流海报内容并通过 LinkedIn 建立联络。 ## Notable quotes > [00:04](https://www.youtube.com/watch?v=qivYz1lEuJE&t=4s) 大家都在使用、部署甚至规模化扩展 AI，然而几乎没有组织真正评估过自身是否在合规、风险与管理层面做好了准备。 — George Antoniou > [02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) 究竟是谁批准了这项工具？谁给出的授权？它走过了什么审批流程？ — George Antoniou > [02:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=151s) 当被问及‘谁在对组织内的 AI 负责’时，几乎没有哪家机构能立刻给出一个明确的名字，也没有设立具备实质权力的常设治理委员会。 — George Antoniou > [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) You cannot govern what you have not prepared for. — George Antoniou > [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) 你无法治理你未曾做好准备的事物。 — George Antoniou > [08:31](https://www.youtube.com/watch?v=qivYz1lEuJE&t=511s) 学校哪条政策规定不能用 AI 写论文？ — George Antoniou > [14:10](https://www.youtube.com/watch?v=qivYz1lEuJE&t=850s) 真正会在未来五年内被取代的，是那些‘不会使用 AI 的人’。AI 本身不会夺走任何人的工作，工作岗位的本质是在经历结构性转移。 — George Antoniou > [15:43](https://www.youtube.com/watch?v=qivYz1lEuJE&t=943s) 这绝不是派领导参加一次研讨会或工作坊就能解决的，而是一个日复一日的持续学习过程。 — George Antoniou > [15:43](https://www.youtube.com/watch?v=qivYz1lEuJE&t=943s) 如果不进行再培训，员工就会输给那些熟练掌握 AI 工具的竞争对手。 — George Antoniou > [18:27](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1107s) AI adoption is a technology decision... AI readiness is a leadership decision... and responsible AI transformation is an organizational commitment. — George Antoniou > [18:27](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1107s) AI 采纳是一项技术决策。AI 就绪是一项领导力决策。负责任的 AI 转型是一项组织承诺。 — George Antoniou > [18:27](https://www.youtube.com/watch?v=qivYz1lEuJE&t=1107s) 你们准备好了吗？ — George Antoniou ## Mentioned George Antoniou, 林恩大学 (Lynn University), 富布赖特项目 (Fulbright Program), 地拉那大学 (University of Tirana), NIST AI RMF, 《萨班斯-奥克斯利法案》 (Sarbanes-Oxley Act), 纽约证券交易所 (NYSE), Waymo, Ai4, LinkedIn ## Video description George Antoniou, Full Professor Cybersecurity, Lynn University Organizations are adopting AI fast. But are they ready to govern it responsibly? This session examines the gap in AI adoption and organizational readiness, drawing on my Fulbright Scholar experience developing AI and cybersecurity initiatives with academia, government, and industry. I will introduce the AI Readiness Framework, a practical model for governance, ethics, leadership, and workforce readiness. You will leave with practical strategies to strengthen AI governance and readiness. If your organization is adopting AI faster than it can govern it, this session is for you. [Source](https://www.youtube.com/watch?v=qivYz1lEuJE)

Source quotes

[0:04] George Antoniou: 大家都在使用、部署甚至规模化扩展 AI，然而几乎没有组织真正评估过自身是否在合规、风险与管理层面做好了准备。 [2:31] George Antoniou: 究竟是谁批准了这项工具？谁给出的授权？它走过了什么审批流程？ [2:31] George Antoniou: 当被问及‘谁在对组织内的 AI 负责’时，几乎没有哪家机构能立刻给出一个明确的名字，也没有设立具备实质权力的常设治理委员会。 [8:31] George Antoniou: You cannot govern what you have not prepared for. [8:31] George Antoniou: 你无法治理你未曾做好准备的事物。 [8:31] George Antoniou: 学校哪条政策规定不能用 AI 写论文？ [14:10] George Antoniou: 真正会在未来五年内被取代的，是那些‘不会使用 AI 的人’。AI 本身不会夺走任何人的工作，工作岗位的本质是在经历结构性转移。 [15:43] George Antoniou: 这绝不是派领导参加一次研讨会或工作坊就能解决的，而是一个日复一日的持续学习过程。 [15:43] George Antoniou: 如果不进行再培训，员工就会输给那些熟练掌握 AI 工具的竞争对手。 [18:27] George Antoniou: AI adoption is a technology decision... AI readiness is a leadership decision... and responsible AI transformation is an organizational commitment. [18:27] George Antoniou: AI 采纳是一项技术决策。AI 就绪是一项领导力决策。负责任的 AI 转型是一项组织承诺。 [18:27] George Antoniou: 你们准备好了吗？

Source quotes

[0:00] George Antoniou: If your organization is adopting AI faster than it can govern it, this session is for you. [8:31] George Antoniou: You cannot govern what you have not prepared for. [11:46] George Antoniou: AI is not going to take anybody's job. The jobs are going to be shifted... [11:46] George Antoniou: Who is going to be replaced by AI is people that they don't know how to use AI in five years. [11:46] George Antoniou: human-in-the-loop [11:46] George Antoniou: human-in-control [18:31] George Antoniou: AI adoption is a technology decision. [18:31] George Antoniou: AI readiness is a leadership decision. [18:31] George Antoniou: Responsible AI transformation is an organizational commitment.

Original transcript

[0:04] Good afternoon. I hope uh everybody's ready  uh for an exciting presentation. My name is [0:10] uh George Antonio. I'm a full professor of cyber  security uh management at league university, [0:16] a full price scholar and recently a fullbride  ambassador. Um I did my um my fullbrite [0:25] um uh scholar residency at the University of Tyona  in Albania in 2025 and this is the fruit of the [0:35] five months uh uh research uh grant I did in that  university. So you will see this is a framework [0:44] that is not based on literature review but an  actual implementation on in the field you know. So [0:51] um this took two universities, two countries,  5 months of work uh to come up with this uh [0:59] AI readiness um and AI adoption as we know  everybody's using AI nowadays. But are we ready [1:09] for uh a compliance, a risk and a management?  Okay. So, um most of you you're probably [1:20] implementing AI in your organizations. Some of you  might using it. Some of you might scaling it. But [1:29] how many of you you assess your organization that  they are ready for adoption or how many of you [1:36] actually went through a formal process to make  a decision how you going to use this technology [1:43] because this is an AI technology it's just a  tool everybody's using it right but what was [1:50] the process that you use to be ready for it did  you go through a governance did you go through [1:57] um uh uh some type of a uh of a process to come  up to a decision. So this is what we're going [2:06] to be talking about. You know, we're going to  be talking between adoption and the speed of [2:11] readiness because this as we know is moving very  very fast. Okay. So this is what it's going to be [2:21] the session that is all about. It's going to be  uh are we are we ready? Are we ready for the AI? [2:31] Okay. So, what is the what is the problem? Okay.  So, the problem is that everybody is adopting AI. [2:44] It's moving very fast. Some of some of us we are  trying to catch up with it. Okay. We're walking [2:52] into it. Some of us were crawling. Okay. But what  is the actual issue behind this technology? Is it [2:59] the technology problem or is it something else  or is this an organizational problem? Well, in [3:06] um nearly every environment uh I work with um  with the universities, the governments, agencies, [3:13] uh it came out through conversations that  someone went and just uh um got this tool. [3:21] Okay. They start using it and they found it  useful. They gave it to other people. They start [3:28] using it and the faculty start using it. Then  the students start using it. Okay. Similar to the [3:35] organizations employees start using it. Okay. They  give it to the other employees. But in reality, [3:40] who approved that? Who gave authorization? They  went through a process. So think about it. Are we [3:49] ready for it for compliance? What is our risk? We  all doing GR GRC governance risk and compliance. [3:57] Most of us are regulated. Okay. So, are we ready  for this? Who assesses the AI? Who pays for it? [4:04] How many tools we're using in the organization?  Who is using it? All these questions, all these [4:11] questions came into a tradition when I started  talking to faculty and we started talking to [4:18] the administration and we start talking to the  students. Okay? Because some institutions they [4:25] were not ready for this to be adopted and  to this day some of them they are still not [4:32] ready. They're not prepared. Are we prepared?  Are you prepared for it? Are you ready for it? [4:38] And when I mean ready for read, you know, are you  ready to accept the risk? Are you in compliance? [4:47] The answer is to be determined, right? When that  you see the compliance comes from a regulation [4:53] and you're going to have to go through an audit or  through an assessment and you're going to realize [5:00] that, oh my goodness, we have not think about  governance. And what do you mean by governance [5:10] is to put some things in auto, right? Otherwise,  they're going to be free for right now AI adoption [5:16] is being used by everybody in the whole globe.  You in the households, in the in in the stores, [5:25] in the taxi drivers, you know, everywhere you go.  Okay, we adopting it in the organizations. How we [5:32] are going to regulate this? What kind of process  do we have? All these questions came into me to [5:40] realize that we have we have a pro uh we have an  issue. We have a gap. Okay. Uh NIS standard you [5:48] know the national is of standards and technology  uh that is run by United States government has an [5:54] AI RM RMF risk management framework. Most of  the most of the organizations says we follow [6:02] that framework but they missing a particular  part. They're not ready for it. Do they have a [6:09] governance? Do they have do they have some type of  policies? Some of them will say yes we do you know [6:15] but if I ask you who's responsible for AI your  organization are you ready to give me a name? [6:23] Do you have a government do you have a governance  committee? Not a not a task force. Do you have a [6:28] governance AI committee? I don't think so. So all  these things you know that they are uh came came [6:37] through uh the realization that everything that  we have out there today points it to that we're [6:46] not ready. We're using it and we're not ready  for it. And when I'm saying we're not ready, [6:52] we're not prepared for it. And it's not that it's  not good. Okay? It's not that it's not bad. We [6:59] need to take action. So this framework actually  is going to help you to be ready for it when the [7:06] time comes to go in and do uh governance risk and  compliance. The last one is the compliance. So the [7:14] the last one is the compliance. The most important  thing in any organization should be compliance. [7:17] What do you mean by that? You know if you prince  Oxley New York real stock exchange public company [7:22] your CEO your CFO and your CEO they're responsible  for what? For making sure you know the integrity [7:30] of data and uh and the financial statements that  they put out there are solid. Ask them about AI. [7:40] Ask them about AI. Are they prepared for it to  testify that their employees are using a specific [7:50] tool or an approved tool in the company? You'll be  surprised of the majority of the tools that each [8:00] organization is using. There is no constant usage.  There is not an approval process and that's the [8:06] issue. So when when I say the issue the gap is to  come up to uh an area that we will be able to give [8:18] ourselves the readiness to be prepared. Bless you.  Okay. So what does that mean? What does that mean? [8:31] There are five dimensions in this framework. Okay.  So I hope you can see that the middle one is the [8:41] uh AI uh readiness framework and then we  have the um interbenders from each other. [8:48] If you start close why close w clockwise uh  you will see the dimensions they are not in [8:55] they're independent from each from each other.  You can do it at any given time but all these [9:01] dimensions all these ingredients that's what  they make the framework. Okay. And this is to [9:06] be ready. So you need to understand what is these  five interconnected dimensions are. Okay. Um the [9:16] central idea is simple. You cannot govern what  you have not prepared for. Does that make sense? [9:25] You cannot govern what you're not prepared for.  If you're not prepared for AI adoption and you [9:31] have many tools in your organization, nobody  authorize it. How you going to govern that? [9:36] So you need to put some governance committee  not a task force to make sure you know that [9:41] these all get approved through a process.  Okay? And this is not about who's using [9:48] it. I just want to make clear. It's about who  owns the decision, who is who is responsible, [9:54] who is accountable. Who who's accountable in your  company now about AI? You have a person who's [10:02] accountable if something goes haywire. I don't  think so. You're going to go down to employee, [10:08] right? And you're going to say, you know, why are  you using this? And then police going to say, what [10:12] is the policy? Whoops, we don't have a policy,  right? Same thing in academia. You you you catch a [10:19] student using AI to do their paper and you report  them and then the student challenge you and say, [10:25] you know, what is the policy? It says not to use  AI, you know, to do my paper. and you go to the [10:29] administration and say do we have a policy about  usage of AI and they say no we have policy about [10:33] plagiaris which is totally different right and  then you debate if the plagiarist tool you know [10:40] identifying the paper is AI use and the and and  the students that uh I wrote the paper the tool [10:46] is wrong and now you go in a debate between  student academia and then your administration [10:54] tries to find out we don't have a policy oh yes we  don't have a policy We because how did you adopt [11:00] this tool? Who approved it? You know, all these  questions that come to place you see my point [11:04] the point is they're not ready. We're not ready.  We need to put this frame framework together. So [11:10] the next ingredient, okay, is is um ethics. But  before I go there, just make sure you know that [11:20] you need to have some type of uh uh an authority  uh in in in the in this in in this dimension. [11:28] Okay. Uh readiness means accountability and  that's the message that I wanted I want to tell [11:34] you about. Okay. And also we need to establish  an I an AI governance committee not a task force. [11:43] Okay. Because a standing committee needs to have  authority. Okay. All right. The ethical part. [11:52] The ethical part is doing what is transparent. Is  it ethical to use AI to write your paper? Is it a [12:01] ethical to have a a faculty grade that paper?  So what is transparency in ethics? And how we [12:09] going to use this for ethical decisions? Take it  outside of academia. You're in healthcare. You [12:16] are a physician. So you make a decision on that.  That's where human in the loop it comes to place. [12:21] And I'm going to talk about that. All right. So,  ethics is another dimension that you need to take [12:26] in consideration when you are getting ready to  be in compliance to address your risks because [12:33] you need to have risks and then you need to have  this governance to govern all these ingredients. [12:39] Okay. So, the other thing the other dimension is  leadership preparedness or readiness. All right. [12:51] Are they prepared? Do they understand what is  AI is all about? And if they do understand, [13:00] do they want to you I want to use AI but I don't  get it. So what do you need to do to prepare that [13:06] to get ready? You need training. Okay. And it's  not just a workshop. It just to train them. All [13:11] right. You need to understand what is this role  in your organization. Okay? It's not just just [13:18] about financial. Okay? It's about policies, right?  We're talking about tokens and spending money and [13:24] all these things. That's not the issue. The issue  is to get these individuals to be ready for that. [13:31] Okay. The next one is curriculum readiness. Is  your in academia who have courses? You know, [13:38] we have curriculums. Are they modernized?  Are they become modern to this AI era [13:45] similar to your company's policy? your standard  operations uh procedures. Okay. Are they are they [13:53] AI embedded into the system? Okay. So that's  that's the other other ingredient to get you [14:01] um ready. Okay. So modernization is alignment with  workforce expectations. So what is your users, [14:10] what your employees, what is your faculty, what  your students, what your customers want out of it. [14:15] The next ingredients is workforce readiness. If  you can see there uh this is not about and most [14:23] of people they worry about this. This is not about  replacing people with AI. And I tell everybody you [14:31] know who is going to be uh replaced with a by  with AI is people that they don't know how to [14:39] use AI in five years. These are the people that  they're going to be replaced. AI is not going [14:44] to take anybody's job. The jobs are going to be  shifted. And if you go back in history, if you go [14:49] back in Manhattan, we used to have carriages, you  know, people that they were sweeping, you know, [14:54] uh the streets. Now we have taxis, right? We  have taxi drivers. Well, what we have now, [14:59] we have autonomous cars, you know, we have way  more, you know, so the the drivers, they not going [15:03] to be there anymore, right? But who's going to be  there? You know, behind each Whimo car, you know, [15:08] there is a a a a control room that there is a  person there who monitors these cars. You know, [15:14] they're not by themselves driving. You know, don't  be mistaken. There's someone who is ready to take [15:19] over if the car, you know, has a manufacturer  or something. That's a human in the loop, [15:24] human in control. Okay, that's a good example. So,  you see the shifting, right? The workforce. So, [15:31] you need to understand the reskilling of your  employees. The next thing, the next thing is to um [15:43] make sure that all these five questions  there, they're being answered. First one, [15:50] you need to know who owns AI. I'll give  you the answer. Your governance committee, [16:00] that's that's your that that's that's the one who  owns AI. You need to have a governance committee [16:06] and that's who owns AI your organization. Okay.  What principles guide guide us. Okay. That's the [16:14] other question. Use ethics. Right. Ethics is what?  It's doing good. Okay. It's what what that's what [16:24] is ethics about. All right. Doing the right thing.  Be transparent. Are your leaders prepared? What's [16:33] the answer there? You need executive education.  And this is not just sending them to a workshop [16:40] or a seminar. This is a continuing. It's a  continuation. These things they evolve day by day. [16:47] AI tools they just growing you know day by day.  Tomorrow there will be a new version, a new model, [16:54] a new thing you know coming out. Your leadership  needs to be trained all the time along with your [17:00] employees. The other question is your workforce  ready and that's where it comes AI literacy and [17:13] reskilling. These are the two ingredients that  you need to understand. Okay, AI literacy means [17:19] to understand what are these tools they do and  reskill it. You need to reskill your employees. [17:25] If you don't reskill it, they will lose their  jobs. Seriously, they will to lose the job to [17:31] someone who knows how to they use these AI tools.  Not because AI taking over, but because someone is [17:38] using the AI tool is curriculum align or your uh  standard operation procedures align. You need to [17:46] moderize your curriculum. If you're in academia,  I motoriize your standards operation procedures [17:50] to embed the AI. All right? So, and when you  get all these right, all these five questions, [17:58] okay, and answers, you have better decisions.  You're going to reduce your risk. You're going [18:06] to create a better trust. You're going to have a a  workforce that is much more confidence. And you're [18:15] going to have an organization that is responsible  for innovation. Because if you not embrace this [18:22] which were you already embracing you know you're  already adopting AI but what is missing is missing [18:26] this frame framework this readiness okay so I  want to leave you with these three statements [18:37] AI adoption is a technology decision okay every  organization I believe in this room has made one [18:47] or is about to make one and what I mean by AI  technology decision is what use what AI tool [18:53] you're going to be using if it's one or multiple  AI readiness it's a literacy is is a leadership [19:02] decision it's not the employees decision it's a  leadership decision if the organization is ready [19:08] or not is based on leadership direction  okay and responsible AI transformation [19:18] is an organizational commitment. So if your  organization is commitment to embrace AI then [19:25] you will have all these ingredients. Okay. So this  starts with uh only you can answer. Are you ready? [19:35] You you can only answer that question not me. If  you're ready or not then use the framework. So [19:42] um this is what I have. I welcome the  opportunity for further conversation. [19:46] I have a poster out there. My Lindic  contact is also on my poster and there [19:52] is no Q&A but I'll be glad to continue after  uh the presentation. Thank you very much.