Washington’s focus on artificial intelligence (AI) includes competing priorities: accelerating innovation, protecting communities, strengthening national security and preparing workers. The White House’s March 2026 legislative framework asks Congress to address these issues through a national approach. Recommendations, however, are not enacted legislation. Washington still has to turn its intentions into requirements. The Economist brought this topic to us in their article, “The debate over AI has taken over Washington.”

Should organizations prepare anyway? And how? Waiting for political certainty seems like an excellent strategy for being spectacularly unprepared.

Let’s start by identifying where AI is already being used, including tools employees adopted without formal approval. Document what information enters those systems, who can access it and which decisions depend on their output. Assign responsibility for monitoring policy developments and translating actual requirements into operational changes.

Most critically, strengthen data governance. Establish clear ownership, consistent definitions, access controls, quality checks and records showing where information originated and when it was updated. AI drawing on outdated policies, contradictory records or poorly labeled information can produce polished nonsense with impressive confidence.

Good governance improves the foundation; it does not guarantee accurate answers. Organizations must also test outputs, check sources and require knowledgeable human review when consequences matter. NIST’s voluntary AI Risk Management Framework provides a useful starting point for organizing that work.

Train employees to recognize limitations, protect sensitive information and report errors. Review vendor terms and establish procedures for correcting failures.

Whether Washington delivers sweeping legislation, narrower rules or another round of announcements, these preparations retain their value. Reliable AI requires reliable information and accountable people.

Keeping data safe and whole is important. Making data accessible is something we know a little about. Whatever you are searching for, it is important to have a comprehensive search feature and quality indexing against a standards-based taxonomy.

Stay current on the ideas shaping data and AI. Subscribe to Taxodiary for insights from the Access Innovations team.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, where smarter AI starts with structured, meaningful, well-governed knowledge.

Artificial intelligence (AI) can draft an article, analyze data and produce an answer before we finish our coffee. But generating something that sounds thoughtful does not automatically make it accurate, meaningful or even wise. The Scholarly Kitchen brought this topic to our attention in their article, “The Representation Is Not the Scholarship.”

“Artificial thought” is not a universally established technical category, however it describes the appearance of thinking: fluent explanations, convincing arguments and conclusions that resemble informed judgment. AI refers to the computational systems performing tasks such as prediction, pattern recognition and language generation. The distinction matters because an impressive performance can encourage us to assume understanding that the output itself does not demonstrate.

Generative AI can produce false information with remarkable confidence. Apparently, being confidently wrong is no longer an exclusively human talent.

Human expertise remains essential because accuracy requires more than polished sentences. A historian recognizes when a quotation clashes with its era. An editor notices when a technically correct statement misrepresents a community. A communications professional understands why the same message can reassure one audience and alienate another.

Context also includes what is missing: whose perspective was excluded, which assumptions shaped the question and whether information applies to this particular situation. These judgments require subject knowledge, attention and accountability.

Effective AI use therefore begins with informed questions and continues with verification. Experts should check sources, test conclusions, identify uncertainty and evaluate consequences before treating generated material as reliable.

AI can expand our capacity to explore ideas and organize information. Its usefulness grows when people bring expertise to the process. The goal should be stronger human judgment, supported by technology, with someone responsible for deciding whether an answer deserves trust.

While AI has undoubtedly enhanced many things, including the way we search, there’s something even more effective at delivering precise, meaningful results—a custom taxonomy. 

Want more insights like this? Subscribe to Taxodiary for practical perspectives on data, AI, taxonomy, and information management.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, the intelligence and the technology behind world-class explainable AI solutions.

Artificial intelligence (AI) is giving fraud a disturbing upgrade. Scammers can impersonate familiar voices, manufacture convincing images and produce polished messages that make yesterday’s suspicious emails look almost quaint. Apparently, even criminals have discovered the benefits of better editing. This important topic came to us from Thomson Reuters in their article, “Deception in the AI Age 2026.”

AI does not invent the desire to deceive. It makes deception easier to produce, personalize and distribute.

Voice cloning illustrates the danger. A caller who sounds like a frightened relative can exploit trust and urgency to demand money. The Federal Trade Commission has warned about these schemes, while the Financial Crimes Enforcement Network has identified fraud involving AI-generated identity documents and other deepfake media.

Yet AI also helps organizations fight fraud. Machine learning can analyze transactions, identify unusual patterns and prioritize suspicious activity for investigation. The US Treasury reported that using machine learning to accelerate check-fraud identification contributed to recovering $1 billion during fiscal year 2024. That is a substantial reminder that the technology’s value depends partly on who uses it and how.

Detection, however, is not a verdict. An unusual transaction may be legitimate, and a convincing fake may escape scrutiny. Organizations should test detection systems, monitor errors, protect personal information and provide human review before consequential decisions.

AI can strengthen both the scam and the response. Preventing fraud therefore requires more than purchasing smarter software. It requires clear procedures, accountable people and permission to question a convincing story. Especially when that story urgently requests your bank details.

Get the latest Taxodiary insights delivered directly to you. Subscribe and stay connected to our experts.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, uniquely positioned to help you in your AI journey.

Artificial intelligence (AI) can write reports, summarize meetings and answer questions with impressive confidence. What it cannot reliably do on its own is determine exactly what your organization means by every word it uses. When “partner” means five different things across three departments, even sophisticated AI has a problem. It may simply deliver that problem in beautifully polished prose.

This is why ontology becomes more important as AI use increases.

In information management, an ontology explicitly defines the concepts within a particular domain, their characteristics and their relationships. A taxonomy organizes information into categories; an ontology goes further by describing how things connect. A program serves an audience, receives funding, advances a goal and operates under particular policies. Those relationships give information usable meaning.

When an organization that uses “member,” “donor,” and “volunteer” throughout its records, one person might occupy all three roles, but those roles carry different expectations, permissions and histories. An ontology can distinguish the person from the roles and clarify how each relates to organizational activities. Without those distinctions, an AI system could retrieve relevant language while assembling a misleading picture.

The stakes grow when AI moves beyond drafting content into recommending decisions or initiating actions. An ambiguous definition that once caused an awkward meeting could influence thousands of automated responses. Greater speed makes shared meaning more consequential.

Ontologies also provide a foundation for connecting information across systems. Departments often describe the same things differently, while identical labels sometimes describe different things. Establishing explicit definitions and relationships gives developers a basis for mapping those differences. The argument is that this work becomes increasingly valuable as organizations ask AI to draw answers from multiple sources.

When incorporated into retrieval and validation processes, an ontology can help a system locate appropriate information and check whether certain relationships make sense. It does not automatically eliminate hallucinations. Generative AI can still produce false claims, a risk identified by the National Institute of Standards and Technology. A structured model of meaning must be paired with reliable sources, testing and human oversight.

There is also an accountability benefit. Explicit definitions give people something concrete to inspect and challenge. What counts as an active participant? Which activities qualify as service? Who decided? An ontology makes those assumptions available for discussion rather than leaving them buried in disconnected documents or individual interpretations.

Of course, an ontology can preserve outdated assumptions just as efficiently as useful ones. It requires stewardship: subject experts, clear ownership, regular review and room for disagreement. AI can assist with proposing concepts and connections, but people must decide whether those proposals accurately represent their work.

Organizations do not need to model everything before using AI. They can begin with a consequential question, define the concepts needed to answer it and expand as useful patterns emerge.

As AI becomes more capable, the responsibility to define meaning becomes more urgent. Before asking machines to connect the dots, we should agree on what the dots actually represent.

The future of AI depends on how content is prepared today. Access Innovations partners with organizations to turn metadata, semantics and structure into AI-ready infrastructure that protects meaning and enables confident innovation.

Go beyond the buzzwords. Subscribe to Taxodiary for informed perspectives on data, AI, taxonomy and the technologies changing how information works.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, uniquely positioned to help you in your AI journey.

As a cinephile, I appreciate a horror mash-up: haunted houses, ancient curses and people who apparently cannot recognize that the basement is a terrible idea. Add taxonomy to the mix and suddenly my love of films meets my appreciation for organized information. Truly, a monster worth creating. This fun but important topic came to us from Darker Times in their article, “Spooky Procedurals: A Taxonomy.”

A taxonomy of Halloween and horror films starts with recognizable categories: supernatural horror, slashers, psychological horror, creature features and horror comedy. But those categories alone cannot capture everything viewers want to find.

Someone seeking a cozy Halloween movie probably does not want an evening of graphic gore. Someone craving vampires may prefer gothic romance over a bloody survival story. Useful classification therefore needs additional dimensions: themes, tone, audience, setting and intensity.

This is where the fun becomes serious. A film can belong to several categories simultaneously. A haunted-house comedy should be discoverable through both its supernatural subject and its comic tone. Forcing it into one exclusive box limits the paths audiences can take to find it.

Clear definitions also matter. What distinguishes psychological horror from supernatural horror? How should overlapping terms connect? Consistent labels, relationships and synonyms help bridge the gap between the language people search with and the language a catalog uses.

Without that structure, discoverability becomes a guessing game. Excellent content disappears behind inconsistent descriptions, vague tags or categories that reflect an organizer’s preferences rather than a viewer’s needs.

Taxonomies do more than tidy shelves. They make meaningful connections between content and intent, supporting browsing, search and recommendations.

Whether organizing research, business information or movies featuring articulate demons, the principle remains the same: people need reliable ways to find what matters. Even monsters deserve decent metadata.

Data Harmony is a fully customizable suite of software products designed to maximize precise and efficient information management and retrieval. Our suite includes tools for taxonomy and thesaurus construction, machine aided indexing, database management, information retrieval and explainable artificial intelligence.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, uniquely positioned to help you in your AI journey.

Artificial intelligence (AI) can predict demand, flag suspicious transactions and recommend actions in seconds. But speed becomes a liability when nobody can explain why a system produced a particular result. “The algorithm said so” is a weak foundation for organizational trust. This interesting subject came to us from MIT News in their article, “Improving AI models’ ability to explain their predictions.”

Explainable AI helps people understand the factors influencing a prediction, the evidence supporting a result and the limitations surrounding both. NIST identifies four essential principles: providing explanations, making them meaningful to their audience, ensuring they accurately reflect the system and recognizing knowledge limits.

That distinction matters. A polished explanation is not necessarily a faithful one. AI can generate convincing language without accurately describing how an answer emerged. Organizations must evaluate explanations alongside predictions, rather than mistake confidence for credibility.

Useful explanations allow employees to question recommendations, identify questionable inputs and investigate potential bias. They also help people affected by automated decisions understand outcomes and raise concerns. An explanation should give someone a practical way to assess a result, not simply another paragraph of technical fog.

Explainability also supports reliability over time. When conditions change, teams need to investigate whether a model still performs appropriately. Understanding influential factors can help them detect problems, adjust processes or decide that human review is necessary.

However, explanation alone cannot guarantee accuracy, fairness or safety. It must operate alongside testing, monitoring and clear accountability.

This is what makes explainable AI essential to sustainability. Responsible adoption requires more than a successful launch; it requires systems organizations can maintain, challenge and improve.

AI earns lasting trust when people can examine its results and act on what they learn. Even impressive technology should be able to withstand a reasonable “Why?”

The future of AI depends on how content is prepared today. Access Innovations partners with organizations to turn metadata, semantics and structure into AI-ready infrastructure that protects meaning and enables confident innovation.

Stay current on the ideas shaping data and AI. Subscribe to Taxodiary for insights from the Access Innovations team.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, the intelligence and the technology behind world-class explainable AI solutions.

Pope Leo XIV’s appeal to protect human art raises a question that reaches beyond museums: What makes an expression authentic? Tech Crunch brought this interesting news to our attention in their article, “Pope Leo XIV is not a fan of AI-generated art.“

That distinction invites reflection on the relationship between creation and experience. A painting shaped by grief, a hymn born of struggle or a poem wrestling with God carries more than an arrangement of colors or words. It represents someone’s encounter with being alive.

Authenticity, however, does not mean freedom from influence. Human artists draw from traditions, teachers, communities and other creators. Faith develops within similar spheres of influence: family, scripture, worship, culture and personal experience. The question is how those influences are absorbed, challenged and made meaningful through a person’s choices.

Artificial intelligence (AI) complicates this relationship. It can produce language about suffering without having suffered or generate religious imagery without practicing faith. Its output may move a viewer, but that response does not establish spiritual experience within the machine.

There is nevertheless room for AI in religious life. Used thoughtfully, it can assist with translation, accessibility, research and creative exploration. A person might use it to develop an idea while remaining responsible for the message, its accuracy and its consequences.

The boundary deserves attention when assistance becomes authority. Fluent answers should not automatically become theological wisdom and convenience should not displace relationships or discernment.

From this perspective, faith offers a framework for evaluating AI: Does its use honor human dignity, acknowledge the people whose work informs it and deepen understanding?

Technology can help shape an expression. Humans must answer for what it means.

If this got you thinking, there’s more where it came from. Subscribe to Taxodiary for the latest insights from our experts.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, where smarter AI starts with structured, meaningful, well-governed knowledge.

Trust becomes difficult when artificial intelligence (AI) delivers answers without showing why. Employees may question recommendations, customers may challenge decisions and leaders may struggle to defend outcomes. Explainable AI helps address that uncertainty by making the factors influencing a result easier to understand. MeriTalk brought this interesting and important topic to us in their article, “Closing the Data and Trust Gaps with Explainable AI.”

Rather than presenting only a prediction or score, an explainable system can identify which inputs influenced its output. For example, a model forecasting inventory shortages might highlight seasonal demand, supplier delays and purchasing patterns. Staff can then assess whether those signals reflect actual conditions.

This visibility can also strengthen confidence in data. When explanations reveal unexpected influences, teams have a starting point for investigating incomplete records, outdated information or misleading relationships. Explainability does not certify data quality, but it can expose reasons to question it and guide improvements.

The benefits extend to organizational trust. Organizations that explain how AI informs decisions give people a clearer basis for asking questions, identifying mistakes and requesting human review. That openness supports accountability, especially when decisions affect access to services, employment or financial opportunities.

Organizations should pair explainability with data governance, performance testing and clear responsibility for decisions. They should also provide ways to challenge outcomes and correct errors.

The goal is informed trust: understanding when AI deserves confidence and when it needs scrutiny. By making decisions more understandable and open to review, explainable AI can help organizations earn that confidence through evidence and accountability.

Go beyond the buzzwords. Subscribe to Taxodiary for practical insights and expert perspectives on data, AI, taxonomy and the technologies changing how information works.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, uniquely positioned to help you in your AI journey.

A tarot reader turns over a card. An analyst opens a dashboard. Both face the same human question: What happens next and what should I do about it?

The tools appear worlds apart, yet both translate complex, uncertain situations into structured interpretations that can guide decisions. One works through symbolic imagery and human association; the other through historical observations and statistical relationships. Neither eliminates the need for judgment.

In a tarot reading, randomly drawn cards provide a framework for interpretation. Readers connect visual symbols with learned meanings, memory, personal experience and the question being asked. A card suggesting disruption might prompt someone to examine instability they have avoided acknowledging. Its usefulness can lie in the questions it raises rather than any promise of supernatural foresight.

Predictive analytics approaches uncertainty differently. Algorithms analyze historical data to estimate future outcomes, such as customer departures, inventory shortages or equipment failures. These estimates emerge from mathematical models whose performance can be tested against observed results. The dashboard may lack velvet and candles, but it still attracts people hoping tomorrow will explain itself.

The comparison becomes especially revealing at the point of interpretation. A tarot card does not specify a business strategy. Similarly, a risk score does not automatically determine the best response. Practitioners must consider context, competing explanations, consequences and available choices. Even a precise probability can leave considerable ambiguity about what action is justified.

Both systems can become narratives. A reader connects cards into a story about conflict and renewal. An analyst connects indicators into a story about declining demand and changing customer behavior. These narratives make complexity manageable, but coherence alone does not establish truth. A persuasive story can conceal weak assumptions as easily as it clarifies useful patterns.

Here lies the illusion of certainty: both tools can dress uncertainty in the comforting robes of an oracle. Symbols acquire authority through tradition and personal resonance; forecasts acquire it through numbers, charts and technical language. Decimal places can look remarkably like destiny when nobody asks how the model was built.

Nevertheless, the distinction matters. Predictive analytics can demonstrate measurable forecasting skill through validation on unseen data, comparison with simpler methods and ongoing monitoring. Tarot should not be treated as an equivalent source of empirically established predictions. Its reflective value and a model’s predictive accuracy are different claims requiring different evidence.

Responsible use begins by acknowledging those boundaries. For analytics, that means examining data quality, assumptions, uncertainty, bias and changing conditions. For tarot used reflectively, it means treating interpretations as invitations to explore possibilities rather than instructions that override evidence or personal agency.

Ultimately, the strongest connection is our desire to give uncertainty a shape. Structure helps us ask better questions and imagine possible futures. Trouble begins when a useful framework becomes unquestionable authority.

Whether consulting cards or computational models, we should ask what supports the interpretation, what might contradict it, and what remains unknown. Foreseeing the future is an ambitious job description. Making a thoughtful decision with imperfect information is a more realistic one.

Tap into decades of expertise in data, taxonomy, metadata and emerging technologies. Subscribe to Taxodiary.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, where smarter AI starts with structured, meaningful, well-governed knowledge.

The latest artificial intelligence (AI) headlines are shaping more than technology conversations. They are influencing investment, hiring and expectations about economic growth. The result is a powerful mix of opportunity and uncertainty. This interesting topic came to us from Tech Crunch in their article, “The ugly economics of consumer AI.”

On the positive side, AI spending is stimulating demand for equipment and infrastructure. This investment supports suppliers and helps sustain economic activity beyond software companies.

AI also offers businesses opportunities to improve productivity, accelerate research and reduce repetitive work. PwC’s 2026 AI Jobs Barometer found that productivity growth was 40 percent higher at companies most exposed to AI than at those least exposed. The research also associated greater AI exposure with stronger hiring and wage growth, suggesting that automation can expand businesses rather than simply shrink payrolls.

However, those benefits are uneven. Skills requirements are changing much faster in jobs exposed to AI. Workers face pressure to learn new tools, while junior employees increasingly encounter demands for judgment and leadership previously associated with senior positions. Training and access to opportunity therefore matter enormously.

Financial enthusiasm creates another vulnerability. Headlines promising extraordinary returns can encourage investment before businesses demonstrate lasting value.

Taken together, these developments suggest that AI’s economic impact depends on how organizations deploy it and how realistically investors assess it. Businesses should measure results, invest in employees and distinguish useful innovation from expensive excitement. Sustainable prosperity requires more than impressive announcements; it requires benefits that reach workers, customers and communities over time.

The future of AI depends on how content is prepared today. Access Innovations partners with organizations to turn metadata, semantics and structure into AI-ready infrastructure that protects meaning and enables confident innovation.

Data and AI aren’t getting any simpler. We can help. Subscribe to Taxodiary for practical insights, trends and expertise.

Melody K. Smith

Data Harmony is an award-winning semantic suite that leverages explainable AI.

Sponsored by Access Innovations, uniquely positioned to help you in your AI journey.