The Ethical Misconception Most Likely to Cause a Third AI Winter
| Why Business Leaders Must Unlearn the Belief That AI Can Reliably Perform Knowledge Work at Human-Expert Levels Without Significant Human Oversight IntroductionThe history of artificial intelligence contains an important lesson that if organisations choose to ignore it could lead to significant damage. Neither of the first two AI winters occurred because AI was completely useless. Instead, both resulted from a gap between what AI could actually do and what influential stakeholders claimed it could do. In the 1970s, expectations around general problem-solving exceeded reality. During the late 1980s, expert systems were marketed as capable of replicating professional judgment through rules and logic, only to reveal fundamental limitations when exposed to real-world complexity. Dangerously, a similar misconception is emerging today: AI can reliably perform knowledge work at human-expert levels without significant human oversight. This belief is not merely a technical misunderstanding. It is an ethical issue involving responsibility, honesty, fairness, risk management, and professional accountability. Ethical leadership requires truthful communication about capabilities and limitations rather than promoting unrealistic expectations. From an ethical perspective, the danger is clear. When organizations remove human oversight based on exaggerated assumptions about AI capability, they transfer risk to customers, employees, patients, investors, and society. This violates fundamental principles of professional conduct and risk management. The issue is not whether AI is valuable. It clearly is. The issue is whether organizations are deploying AI responsibly and transparently, particularly in domains where errors can create significant harm. Challenges Confusing Fluency with UnderstandingLike their grandmother Elisa, modern AI systems generate responses that appear intelligent, confident, and authoritative. However, convincing language is not the same as genuine understanding. Ron Jeffries, one of the co-creators of Extreme Programming, has repeatedly warned about confusing visible outputs with actual value and understanding. Metrics, demonstrations, and impressive presentations can create an illusion of capability while masking underlying limitations. Ethically, this creates a challenge for leaders. Employees and stakeholders often assume that articulate AI responses indicate expertise. AI systems can produce inaccurate recommendations while sounding completely confident. When leaders accept fluency as proof of competence, they risk making decisions that affect people's livelihoods, finances, health, and safety. Removing Oversight Before Building Verification One of the most troubling trends in the current AI cycle is the movement from assistance toward autonomy. AI initiatives frequently focus on reducing human involvement. Yet risk management practice emphasizes that risk management should be integrated into decision-making processes and that uncertainty must be actively managed rather than ignored. Risk is fundamentally the effect of uncertainty on objectives and must always be taken into consideration when decisions are made. Many organizations are pursuing cost savings through automation while delaying investments in verification, auditing, monitoring, and governance mechanisms. This reverses the logical order of responsible risk management. Ethically, oversight should not be removed because technology appears impressive. Oversight should only be reduced after evidence demonstrates that risk remains within acceptable limits. High-Stakes Domains Magnify Ethical Risk It is interesting to see that the strongest push for autonomous AI is occurring in environments where mistakes matter most:
Errors in these domains carry consequences that extend beyond productivity losses. They may affect patient outcomes, legal rights, financial security, privacy, regulatory compliance, and public trust. PMI's ethical framework emphasizes acting responsibly and protecting stakeholders. Similarly, risk management practices stress proactive management of uncertainty and transparent decision-making. Allowing AI systems to operate with insufficient human review in high-consequence environments creates ethical exposure that organizations may underestimate. Overreliance on Best-Case Demonstrations Vendor demonstrations typically showcase ideal scenarios. Real-world work rarely resembles these controlled conditions. Agile ways of working emphasize continuous feedback, collaboration, transparency, and adaptation to actual operating conditions rather than assumptions. Ethically responsible leaders must recognize that demonstrations are hypotheses, not proof. A technology that performs well in a polished demonstration may behave very differently when exposed to incomplete information, conflicting requirements, organizational politics, regulatory constraints, and ambiguous stakeholder needs. Ignoring the Human Dimension of Knowledge Work Agile practices demonstrate that Agility depends not only on knowledge but also on the capability to interpret, adapt, learn, collaborate, and respond to changing conditions. Knowledge application requires context and judgment. Knowledge work is rarely a simple process of retrieving information. It requires:
Current AI systems can support these activities but cannot reliably replace the full spectrum of human responsibility that accompanies them. RecommendationsMatch Oversight to Consequence Not every AI output requires the same level of review. Low-risk activities such as brainstorming, drafting, or summarization may require limited supervision. High-risk activities involving legal, medical, financial, or strategic decisions require rigorous human validation. This approach aligns with PMI's Code of Ethics principles of responsibility and fairness and follows a risk-based decision-making philosophy. Build Verification Before Autonomy Organizations should establish:
before expanding AI autonomy. Agility is not about eliminating controls. True agility balances learning, adaptation, and accountability. Prioritize Transparency and Honest Communication The PMI Code explicitly highlights honesty as a core professional value. Leaders should avoid overstating AI capabilities to executives, boards, customers, or regulators. Ethical communication means:
Trust grows when organizations communicate reality rather than hype. Treat AI as a Knowledge Amplifier, not a Knowledge Replacement The Agile Manufacturing Enterprise concept, defined in 1991, suggests that organizational success emerges from balancing knowledge management and response capability. Knowledge without appropriate application creates little value. AI should be viewed as:
rather than a wholesale replacement for professional judgment. Establish Ethical AI Governance Organizations should create governance frameworks incorporating:
Such practices align with both PMBOK risk-management principles and emphasis on continual improvement, stakeholder engagement, and integrated governance. The Bottom Line The greatest threat of a third AI winter is not that AI lacks value. It is that organizations may once again confuse genuine capability with exaggerated expectations. The ethical danger lies in believing that AI can reliably perform human-expert knowledge work without meaningful oversight. History shows that such claims can damage far more than individual projects. They can undermine trust in an entire field. The lesson is not to reject AI. The lesson is to deploy it responsibly. Business leaders who embrace this principle will recognize that AI's long-term impact is likely enormous. However, achieving that impact requires honesty about present limitations, disciplined risk management, and unwavering commitment to ethical responsibility. Organizations that combine AI capability with human accountability will create sustainable value. Organizations that pursue autonomy without verification risk repeating the mistakes that contributed to previous AI winters. The future of AI will not be determined solely by technological advancement. It will be determined by whether leaders choose ethical stewardship over short-term optimism. Question for Readers: Should business leaders understand AI capabilities honestly, or should expectations be shaped more by vendor demonstrations than real-world evidence? |
Agile vs Traditional Risk Management debate. An Ethical reflection
IntroductionRisk management is a cornerstone of responsible project delivery, yet the debate over when and how to manage risks remains fierce. Should risks be identified and analysed comprehensively upfront, as in traditional predictive methodologies, or should they be managed continuously throughout delivery, as Agile frameworks propose? This question is not just a matter of methodology; it is fundamentally ethical, touching upon our duties to clients, teams, stakeholders, and society at large. This post explores the ethical dimensions of the Agile vs. Traditional risk management debate. ChallengesUpfront Identification: Governance vs. Uncertainty The traditional predictive approach emphasizes comprehensive risk registers and formal upfront analysis. As outlined in the PMBOK and ISO 31000, this method has the ability to reduce surprises and provide transparency, key for governance and audit requirements. It aligns with PMI’s ethical principle of responsibility—ensuring all foreseeable risks are considered and documented. However, the complexity and dynamism of modern projects mean that many risks cannot be foreseen at the outset. Over-reliance on upfront planning may create a false sense of security and stifle responsiveness. Continuous Management: Adaptation vs. Oversight Agile frameworks advocate for continuous risk identification and adaptation. This approach recognizes that most risks emerge during delivery, especially in complex or innovative projects. Continuous inspection aligns with PMI’s values of honesty and respect—facing risks as they arise, communicating transparently, and adapting ethically. Yet, critics argue that this approach may lack the rigor required for governance, potentially overlooking systemic risks or failing to meet audit standards. Ethical Dilemmas: Transparency, Accountability, and Trust Both approaches present ethical dilemmas. Waterfall’s upfront analysis supports accountability and transparency but may lead to bureaucratic inertia or ignore emerging threats. Agile’s ongoing adaptation fosters trust and openness but could result in missed documentation or gaps in formal oversight. The PMI Code of Ethics emphasizes balancing stakeholder interests, which is challenged by both extremes. Recommendations Blend Approaches for Ethical Integrity Research and best practices suggest that an ethical approach to risk management blends the strengths of both models. Initial identification and documentation should be robust enough to satisfy governance and audit requirements. However, teams must also commit to continuous risk inspection, adaptation, and transparent communication, in line with Agile values and PMI’s ethical standards. Prioritize Stakeholder Engagement Ethical risk management requires ongoing stakeholder engagement. This means not only communicating risks early and often but also ensuring that stakeholders understand the evolving risk landscape. As the PMBOK and Agile Practice Guide emphasize, fostering dialogue and trust is crucial to responsible project delivery. Document Adaptations Transparently To meet both audit and ethical requirements, all risk adaptations should be documented as they occur. This satisfies the PMI’s principles of fairness and honesty and ensures that lessons learned are shared across teams and organizations. Foster an Ethical Culture Project leaders must cultivate a culture where risk is everyone’s responsibility. This means encouraging team members to speak up about emerging risks, rewarding transparency, and never penalizing those who surface inconvenient truths. The Bottom LineThere is no one-size-fits-all answer to the Agile vs. Traditional risk management debate. The ethical path lies in integrating comprehensive upfront analysis with continuous risk management. By respecting governance requirements and embracing adaptive practices, project professionals can fulfill their duties to clients, teams, and society. Ultimately, ethical risk management is about more than compliance—it’s about stewardship, transparency, and the courage to confront uncertainty as it arises. Question for Readers: What ethical challenges have you faced when balancing governance requirements with the realities of Agile project delivery? |
The Hidden Risk of AI in Agile: The Illusion of Velocity
IntroductionArtificial Intelligence is transforming the way Agile teams approach project delivery. With AI-powered tools, teams can generate requirements, code, tests, documentation, and analysis at unprecedented speeds. At first glance, this seems like a breakthrough—velocity metrics soar, and delivery pipelines hum with activity. However, as Agile practitioners know, true success rests not on how much is produced, but on how well teams understand the work, align with customer needs, and continuously learn and adapt. This post explores the biggest risk AI introduces to Agile delivery: the illusion of increased velocity without corresponding gains in understanding, quality, or business value. Challenges: When AI Outpaces UnderstandingAI can turbocharge output in almost every domain of software delivery. User stories, acceptance criteria, test cases, and even working code can be produced in minutes. But when teams lean too heavily on AI, a subtle but dangerous problem emerges. The Mirage of Progress
A Real-World Example Consider a team that uses AI to generate user stories, acceptance criteria, and large swaths of application code. Sprint velocity doubles, and the product appears to advance rapidly. Six months later, the reality sets in:
Maintenance slows to a crawl, and the product’s value to customers diminishes. The illusion of progress gave way to very real problems. Recommendations: Keeping AI as a CopilotAI is a powerful tool—but in Agile, it must be harnessed thoughtfully. Teams can mitigate the risks by reinforcing core Agile values and practices: Treat AI as a Copilot, Not a Decision Maker Use AI to augment team capabilities, not replace human judgment. Critical decisions about requirements, architecture, and quality must remain with the team. Strengthen Accountability Maintain human accountability for all project artifacts. The PMI Code of Ethics underscores the responsibility to act with integrity and transparency. Rigorous Code Reviews and Validation Scrutinize all AI-generated outputs. Peer reviews, pair programming, and automated tests help uncover defects and ensure alignment with business goals. Measure What Matters Focus on customer value, quality, and cycle time—not just output or velocity. Align metrics with desired outcomes, as recommended in the PMBOK and Agile Practice Guide. Foster Shared Understanding Continue Agile ceremonies such as daily stand-ups, sprint planning, reviews, and retrospectives. These rituals build context, encourage feedback, and promote continuous learning. Educate Stakeholders Set realistic expectations about what AI can and cannot do. Stakeholders must understand that faster output does not guarantee better outcomes. Continuous Risk Management Apply principles from ISO 31000 to identify, assess, and manage risks associated with AI-driven delivery. Make risks visible and address them proactively. The Bottom LineThe greatest risk AI introduces to Agile delivery is not poorly written code, but the false confidence that faster output equals better delivery. As Agile professionals, our responsibility is to ensure that increased velocity is matched by deeper understanding, higher quality, and true business value. By treating AI as a powerful assistant—not an infallible expert—and by reinforcing human accountability, we can harness AI’s strengths without falling prey to its risks. Questions for Readers: How do you educate stakeholders about the realities of AI-driven Agile delivery? |
An Ethical Reflection on Using AI in Hiring: Respect, Responsibility, Fairness, and Honesty
IntroductionArtificial Intelligence (AI) is transforming organisations and, unavoidably, is used in the hiring process. From resume screening and skill assessments to behavioural analysis and culture fit evaluations, AI promises speed, consistency, and efficiency. But with this technological leap, ethical questions arise: Do these systems honour the fundamental values of respect, responsibility, fairness, and honesty? I hired hundreds of professionals for over 40 years, as a line manager or as a project manager. I see one of my career achievements as the hiring of over 100 university graduates, people without any practical experience who perhaps won’t pass even the CV screening phase. I never had a checklist or a set of questions. I had an opinion after the first 5 minutes, and time proved that I made the right decision. Most graduates had a notable contribution to the organisation’s success, and they had a successful career. This blog post explores the ethical landscape of AI-driven hiring. ChallengesRespect: The Human Element PMI’s Code of Ethics emphasizes respect for individuals, including privacy, dignity, and autonomy. When AI screens candidates, does it respect the nuances of human experience? AI systems rely on data — often stripped of context. While this can help anonymize candidates and reduce overt bias, it risks overlooking unique backgrounds or non-traditional career paths. AI cannot “read between the lines” in the way humans can; it may miss stories of resilience or innovation not captured in keywords or structured data. Responsibility: Accountability in Automation Responsibility demands clear accountability for decisions and outcomes. When AI makes a hiring decision, who is responsible — the algorithm designer, the organization, or the tool itself? Ambiguity in responsibility can erode trust and create ethical blind spots. If an AI system rejects a qualified candidate due to biased training data or flawed logic, assigning responsibility can be challenging. Organizations must ensure there is always a human-in-the-loop and clear lines of accountability. Fairness: Unintended Bias Fairness for the candidate, team and organisation is a cornerstone of ethical hiring. AI systems, trained on historical data, may perpetuate or even amplify existing biases. For example, if previous hiring patterns favoured certain demographics, AI might “learn” these preferences, disadvantaging underrepresented groups. While AI can reduce some forms of human bias, it can also encode and scale bias at an unprecedented rate. Hiring a candidate that is not a good fit for the team or is not aligned with the organisation’s ethical values and strategic goals is unfair to the team and the organisation as a whole. Transparency in model design and continuous monitoring are critical to mitigate these risks. Honesty: Transparency and Trust Honesty involves openness and truthfulness in communication and process. Candidates should know when and how AI influences their evaluation. Although no longer a standard practice, candidates deserve honest feedback. If AI decides, can it explain why? Many AI models, especially deep learning systems, are “black boxes” with decision-making processes that are hard to interpret. This lack of transparency can undermine trust among candidates and stakeholders. Can AI Read Between the Lines or Assess Team Fit? AI excels at analysing structured data but struggles with the subtleties of human communication and team dynamics. Team fit is nuanced, often involving non-verbal signals, intuition, and shared values — areas where AI still lags behind humans. While AI can assess personality traits or match skills to job descriptions, it cannot fully understand how an individual’s unique qualities will mesh with a team’s culture. The Impact of Bias: Getting the Right Candidate Bias in AI can lead to missed opportunities and reinforce systemic inequities. When AI filters out qualified candidates due to biased data, the organization loses potential talent and diversity. Moreover, if candidates perceive the process as unfair, it can damage the employer brand and erode trust in the system. Fairness is not just about process but about outcomes that reflect ethical intent. Can AI Select Better Than Humans? AI offers consistency and can process vast amounts of data free from fatigue or mood. However, humans bring empathy, intuition, and contextual understanding. The best outcomes may come from a hybrid approach: AI for efficiency and humans for judgment. Balancing AI tools with human oversight may ensure both quality and ethical integrity. RecommendationsImplement Transparent AI Systems: Use explainable AI models and communicate openly with candidates about how AI is used in hiring. Provide clear feedback channels. Ensure Human Oversight: Always involve humans in final hiring decisions. Assign clear responsibility for outcomes. Continuously Audit for Bias: Regularly review AI systems for unintended bias. Use diverse training data and seek input from stakeholders across backgrounds. Prioritize Fairness and Respect: Design AI systems that respect individual differences and accommodate non-traditional candidates. Value diversity and inclusivity in both the hiring process and the resulting teams. Foster Ethical Awareness: Train hiring managers and AI developers on ethical standards. Encourage ongoing learning and ethical reflection. The Bottom LineAI has the potential to be a powerful tool in the hiring process, offering greater efficiency and the potential for more objective decision-making. However, ethical values — respect, responsibility, fairness, and honesty — must remain central. AI cannot yet “read between the lines” or fully understand team dynamics. Bias is a real risk, and unchecked, it can undermine the goal of finding the best candidate. Ethical and effective hiring practices should combine the strengths of AI and human judgment, guided by clear ethical standards and continuous oversight. Question for Readers: Can technology truly replace the human touch in understanding team fit and potential? |
The Ethics of Externalising Risk: Rethinking “Fail Fast” and MVP in Product Development
IntroductionAgile changed the way products and services are developed, aligning processes with the fast and complex changes in the business environment. Nowadays, in the high-velocity world of tech innovation, the “fail fast” mantra and the Minimum Viable Product (MVP) concept have become cornerstones of Agile product development. Teams are encouraged to release early, learn rapidly, and iterate based on real-world feedback. While these approaches can accelerate learning and reduce wasted effort, an emerging ethical dilemma shadows their popularity: What happens when “failing fast” means delivering an unfinished or untested product, exposing real users to privacy violations, security flaws, or even physical harm? Is there a risk that the watermelon effect (Green outside, red inside), used to describe unethical project reporting, can occur in Agile product development? At what point does the drive for rapid feedback cross ethical boundaries, risking harm to users or the public? This blog post explores the ethical conflicts inherent in overusing MVP and “fail fast” strategies, examines the challenges and proposes recommendations for ethically navigating the tension between speed and responsibility. ChallengesThe Allure—and Danger—of “Fail Fast” The “fail fast” mindset encourages experimentation and learning by rapidly deploying new features or products. When aligned with Agile values, this can be a powerful driver of innovation. However, overreliance on MVPs and under-tested releases can externalise risk—transferring the burden of failures from organisations to unwitting users. This externalisation is especially dangerous when it comes to privacy, security, and safety. Complex systems are increasingly interconnected, and the speed of change can outpace the ability to foresee consequences. Releasing features with minimal testing may expose users to data breaches, non-compliance with regulations like GDPR, or even physical harm in cases involving IoT or health tech devices. Ethical Responsibilities and Professional Codes The PMI Code of Ethics and Professional Conduct emphasises responsibility, respect, fairness, and honesty. It requires practitioners to make decisions in the best interests of society, public safety, and the environment. Similarly, Risk Management Standards highlight the importance of integrating risk management into organisational processes, rather than relegating it to an afterthought. Yet, in the race to outpace competitors, teams may deprioritise security, privacy, and compliance in favour of rapid deployment. This creates an ethical conflict: Should organisations prioritise quick market feedback, or their obligation to protect users and comply with legal standards? Ron Jeffries, co-creator of Extreme Programming, cautions that “working software” is not enough—software must also be safe and reliable. The Agile Practice Guide warns against anti-patterns where teams treat user trust as expendable in pursuit of learning. Regulatory and Societal Pressures With the rise of privacy regulations such as GDPR and increasing scrutiny from both regulators and the public, companies are under pressure to demonstrate that they take user safety and privacy seriously. Failing to do so can result in legal penalties, reputational harm, and a loss of trust that is difficult to regain. Recommendations1. Strict Definition of Done (DoD) Adopt a non-negotiable Definition of Done that includes comprehensive security checks, privacy assessments, regulatory compliance, and rigorous testing. As recommended in the Agile Practice Guide and PMBOK®, these criteria should be built into every iteration, not left for later stages. Definition of done should be part of the product Backlog item creation and confirmed with the Product Owner in the Sprint planning. This practice ensures that ethical and legal obligations are met before exposing users to new features. 2. Explicit Risk Backlogs Treat risk mitigation, privacy concerns, and technical debt as first-class citizens in your backlog. Modern risk management thought holds that risks should be transparent and actively managed—not hidden or deferred. By maintaining a visible risk backlog, teams can prioritise and address potential harms alongside business features. Risks and issues must be part of the Product Backlog as separate item types and must be linked with tasks to reduce the impact of issues and negative risks or take advantage of positive risks. 3. Servant Leadership & Protection Leaders must adopt a servant leadership model actively shielding teams from pressures to cut corners for short-term wins. Leaders should foster a culture where ethical risk disclosure is valued over meeting aggressive KPIs and where speaking up about potential harms is encouraged and rewarded. 4. Continuous Ethical Training and Review Integrate ongoing ethics training, drawing from the PMI Code of Ethics, into team routines. Encourage regular review of ethical dilemmas, and provide forums for discussing the trade-offs between speed and responsibility. 5. Stakeholder Engagement and Transparency Engage end-users and stakeholders early and often—not just as test subjects, but as partners in risk identification and mitigation. Transparency about what is being tested, potential risks, and the steps being taken to protect users builds trust and helps organisations identify blind spots. The Bottom LineThe “fail fast” culture and MVP approach, when applied without ethical guardrails, can shift unacceptable risks onto users and society at large. As professionals guided by established codes of ethics and global standards, it is imperative to balance the drive for rapid learning with the obligation to protect user privacy, safety, and compliance. By embedding robust definitions of done, explicit risk management, servant leadership, and continuous ethical reflection into Agile practices, communities can innovate responsibly—delivering value without sacrificing trust. Ultimately, the most sustainable path to innovation is one that honours both speed and stewardship. Question for Reflection: How can your organization ensure that risk management and ethical considerations are not sacrificed for speed? |
