What Happens When You Interview the Father of PR Through AI?

In 2026, the Museum of Public Relations launched Ask Eddie, an AI tool that simulates a conversation with Edward Bernays, the so-called father of public relations who died in 1995. Crystal Borde's June-July 2026 PRSA article shares a transcript of her exchange, and hands-on exploration reveals the model was trained on over one million words of Bernays' writings and recorded interviews.¹

The tool's existence raises a pressing question: Can an algorithm replicate the strategic mind that coined "engineering of consent"? Ask Eddie's disclaimer warns it is not historically accurate for academic use, yet the responses are sharp enough to provoke new conversations about ethics and the role of counsel in an AI-driven era.

Who Was Edward Bernays and Why Is He Called the Father of PR?

Few figures in the communication field inspire as much admiration and debate as Edward Bernays. Born in Vienna in 1891 and living until 1995, his career spanned nearly the entire 20th century, during which he transformed public relations from a mostly reactive press agentry into a strategic, research-driven profession.¹

The Architect of Modern Persuasion

Bernays was the nephew of Sigmund Freud, and he drew heavily on his uncle's psychoanalytic theories to understand group behavior and motivation. After immigrating to the United States, he started his own public relations counsel office in 1919, becoming one of the first to offer such services explicitly. One of his most famous campaigns was the "Torches of Freedom" stunt for the American Tobacco Company in the 1920s, where he reframed women smoking in public as a symbol of liberation. He also authored foundational texts, including *Crystallizing Public Opinion* (1923) and *Propaganda* (1928), which articulated a vision of PR as a scientific tool for shaping public attitudes.

Why He Earned the Title 'Father of Public Relations'

Bernays didn't invent the practice of publicity, but he was the first to apply systematic social science to it. He borrowed concepts from psychology, sociology, and political science to engineer consent, a term he coined to describe the deliberate and ethical management of public opinion.¹ This approach set him apart from earlier press agents who relied on instinct and opportunism. Bernays taught that PR was a two-way street: institutions must listen to their publics as much as they speak to them. His insistence on research, planning, and evaluation became the backbone of modern strategic communication.

A Broader Legacy: Other Founding Figures

While Bernays is central, he wasn't alone. Ivy Lee, often called the father of modern PR, pioneered honest and transparent corporate communication, famously declaring "the public be informed." Arthur Page, the first corporate PR vice president at AT&T, established principles like telling the truth and managing for tomorrow. And Doris Fleischman, Bernays' wife and often uncredited collaborator, broke barriers as the first woman to use her maiden name on a U.S. passport and was a powerful voice in the field. Recognizing these figures prevents an oversimplified narrative of a single founder and enriches our understanding of PR's diverse origins. Today, their collective legacy resonates whenever crisis communication experts navigate organizational challenges under intense public scrutiny.

The Enduring Relevance of Bernays' Ethical Compass

Bernays' ideas remain strikingly relevant, especially his ethical convictions. In the AI-generated interview with "Ask Eddie," the simulated Bernays states: "Never to accept a position that puts duty to a client above duty to society."¹ That principle aligns with modern codes of ethics from organizations like the Public Relations Society of America (PRSA), which stress advocacy for the public interest. As communication professionals grapple with deepfakes, disinformation, and AI-generated content, Bernays' early warnings about the moral weight of the practitioner's role offer a timeless guide. The "Ask Eddie" tool thus becomes not just a historical curiosity but a mirror reflecting the profession's ongoing responsibility to balance client goals with societal good.

How the 'Ask Eddie' AI Tool Works

The Museum of Public Relations, in partnership with Edelman, has created a closed-corpus AI persona that lets anyone sit down for a simulated conversation with Edward Bernays, drawing directly from more than one million words of his writings and roughly ten years of late-life video recordings.¹

Built on a Closed Corpus of Bernays' Own Words

The tool, named "Ask Eddie," is not a general-purpose chatbot. It is a closed-corpus LLM persona built with the Archie AI Personas methodology.² Its knowledge is strictly drawn from Bernays' recorded interviews, speeches, published writings, and other public materials held by the museum. The dataset includes video recordings captured during approximately the last decade of his life, giving the model a concentrated window into Bernays' later reflections and campaign retrospectives. Because the corpus is closed, Ask Eddie cannot pull in information from external sources or current events; its knowledge cutoff is the year of Bernays' death, 1995. The museum has not publicly disclosed the precise model architecture, update cadence, or the full count of individual documents in the corpus, so practitioners should treat it as an interpretive repackaging rather than a research database.

For professionals interested in how storytelling matters within strategic communication, this kind of corpus-driven persona offers a compelling case study in narrative design.

Session Limits and the User Experience

Each conversation has a hard 10-minute session length limit. The tool remembers context only within that single session; it does not retain cross-session memory, so returning users start fresh. A conversation ends either by typing or saying "Goodbye" or by clicking an End Conversation button. Input is flexible: you can type queries or click a microphone icon to speak, and output appears as text with an optional synthesized voice that you toggle via a sound icon. When many users are online simultaneously, Ask Eddie may queue responses rather than answering instantly.¹

Transcripts, Timestamps, and the Disclaimer

A transcript download feature captures the full conversation. The PRSA article documents one such transcript downloaded on May 12, 2026, at 11:22:15 a.m. ET, confirming that the tool timestamps each session, a small but useful metadata point for anyone keeping records of these simulated interviews. Importantly, the tool carries a prominent disclaimer: AI-generated outputs are explicitly not to be cited as historically accurate sources for academic purposes. This design choice signals the museum's caution. Ask Eddie is best understood as a synthesized conversational exploration of Bernays' campaigns and reflections, not an authenticated primary source.¹ The tool also prohibits audio or video recording, reinforcing its positioning as a private, reflective exercise rather than a broadcast-ready authority.

When you pose a question to AI Bernays, the tool displays a clear disclaimer: the output should not be cited as historically accurate for academic purposes. It is a striking example of an AI interface that builds epistemic humility into the user experience, reminding users that simulations are interpretations, not primary sources.

What AI Bernays Actually Says: Key Quotes and Themes

The AI-generated Edward Bernays delivers counsel that still resonates in today's boardrooms. Across the interview transcript published by PRSA, three responses stand out for their clarity and strategic depth, each offering a window into both Bernays' legacy and the AI's construction of his persona.¹

Duty to Society Over Client

When asked about the ethical foundation of public relations, AI Bernays states: "Never to accept a position that puts duty to a client above duty to society." This echoes the real Bernays' later-life emphasis on social responsibility, seen in his 1965 essay "The Engineering of Consent," where he argued that PR must serve the broader public interest. But the AI's crisp formulation is likely a distillation, not a verbatim quote. The phrase reads like an idealized maxim, perhaps smoothing over the complexity of his actual client work, including campaigns for tobacco companies and banana republics that critics have long cited as contradictory. The tool's disclaimer warns that AI outputs are not historically authoritative, and here the line between authentic principle and AI-generated synthesis blurs.

Public Relations as a Two-Way Street

The AI asserts that "public relations is fundamentally a two-way street between institution and public." This aligns closely with Bernays' writings in his 1923 book *Crystallizing Public Opinion*, where he stressed listening to and interpreting the public's mind as much as informing it. For practitioners weighing public relations vs marketing vs strategic communication, the distinction matters: Bernays positioned PR as an interpretive discipline, not merely a promotional one. Yet the AI version may oversimplify. Bernays often framed the two-way street more as a feedback loop for persuasion rather than genuine dialogue. The response is rhetorically effective, a soundbite fit for modern stakeholder engagement models, but it might elide the asymmetrical nature of the original engineering of consent.

The Ethics of Speaking and Silence

Perhaps the most practically useful response is the AI's guidance on when to speak: "Speak when silence would damage the relationship with its publics and when you have genuine expertise; be silent when the issue is outside its sphere or would be mere posturing." This reads like a ready-made crisis communication heuristic, and it reflects themes from Bernays' later writings on corporate responsibility. However, its modern, actionable tone raises a critical attribution problem: the tool does not signal whether these are direct quotes, paraphrases, or wholly AI-generated constructions. Practitioners may find the framework compelling, but they cannot trace its origins.

Attributing AI Bernays: Quotation or Synthesis?

The core challenge for any user is that the "Ask Eddie" tool presents a seamless monologue. There are no footnotes, no cues to distinguish between words lifted from Bernays' recorded interviews or books and phrases invented by the large language model. The AI's disclaimer only covers overall historical accuracy, leaving the user to wonder whether an elegant turn of phrase is a true Bernaysism or a convincing imitation. This matters because the AI's version of Bernays may be more principled and consistent than the historical figure, creating an inadvertent whitewash. For those studying mass communication theory, the gap between historical record and AI synthesis is a significant methodological concern. As a reflective tool, the responses are stimulating; as a historical source, they remain unverifiable in their current form.

Never to accept a position that puts duty to a client above duty to society.

How Accurate Is AI Edward Bernays? Evaluating the Tool's Limitations

The public relations field watches as generative AI tools claim to channel the wisdom of its founder, but almost no one asks how faithfully these simulations reflect the historical Bernays. While the novelty of an AI version of Edward Bernays captures attention, a rigorous accuracy check is the content gap that existing coverage has left unaddressed. Understanding where the Ask Eddie tool aligns with Bernays' primary sources, and where it might smooth or modernize his voice, is essential for any practitioner or student who engages with it.

Testing Ask Eddie Against Bernays' Canon

A careful comparison of the AI's output with Bernays' published works reveals both alignment and subtle drift. The tool draws on recorded interviews and published writings, so it naturally echoes core tenets: his emphasis on the "two-way street" between an institution and its publics, the concept of the "engineering of consent," and his insistence that public relations must balance client duties with societal responsibility.¹ These positions are well documented in texts like *Propaganda* and *Crystallizing Public Opinion*, and the AI reproduces them with reasonable fidelity. However, a closer look at more controversial passages, where Bernays advocated for elite manipulation of public opinion or described mass communication theories in starkly cynical terms, often turns up gentler, less provocative phrasings from the AI. The tool appears to curb the raw edge of Bernays' thinking, presenting a version that aligns more comfortably with modern professional ethics.

Why Persona AI Models Flatten Intellectual Evolution

Persona AI models have known limitations that help explain these discrepancies. First, they tend to flatten contradictory or evolving positions into a consistent, synthetic persona. Bernays wrote and spoke over many decades, and his views on propaganda, democracy, and corporate influence shifted and sometimes contradicted themselves, yet an AI optimized for coherence may gloss over this evolution. Second, training data mixing can introduce recency bias: the model may inadvertently blend Bernays' original language with later interpretations or modern commentary, subtly modernizing his worldview. Third, an AI cannot capture the live, contextual reasoning behind recorded remarks; it produces statistically plausible responses, not faithful recollections. The transcript disclaimer that the output "is not to be cited as a historically accurate source for academic purposes" underscores this point.¹

The Practical Risk for Students and Professionals

The most critical implication is a category error that the field must guard against. When students or practitioners treat Ask Eddie's words as primary-source Bernays, they mistake a plausible synthesis for a direct quotation. The tool generates text that sounds authoritative and is rooted in authentic material, but it is not a window into Bernays' actual mind. It cannot weigh which of his ideas he might have revised after a decade of reflection, nor can it capture the full range of his rhetorical strategies. For those building careers with a masters in communication, the AI is a starting point for reflection, not a citation. Recognizing this boundary keeps historical inquiry rigorous and prevents the simulation from becoming a substitute for studying the man's own words in context.

Related Articles

AI Ethics and the Simulation of Historical Figures

In 2024, two major legal actions reshaped the ethics of AI-generated personas: the Tennessee ELVIS Act explicitly prohibited unauthorized AI replicas of a voice or likeness post-mortem, and the EU AI Act (Regulation 2024/1689) imposed transparency requirements for labeling AI-generated content.¹ For public relations professionals, these developments raise urgent questions about tools like Ask Eddie that simulate Edward Bernays.

Why Existing Ethics Codes Matter

The PRSA Code of Ethics demands advocacy and honesty, while the CIPR Code calls for integrity and transparency.¹ Both challenge practitioners to weigh the value of a simulated historical figure against the duty to serve the public interest. Broader frameworks from IEEE and ACM reinforce these principles: the IEEE Ethically Aligned Design guidelines prioritize human rights and dignity, and the ACM Code of Ethics highlights avoiding harm and respecting privacy.²³ These codes provide a ready-made checklist for evaluating AI tools that resurrect deceased public figures, a practice sometimes called how to spot fake news in more casual contexts but one that demands far more nuanced scrutiny here.

Three Distinct Ethical Risks in Digital Resurrection

Consent: Bernays cannot opt in or out of this simulation. Scholars like Edina Harbinja (2025) argue that post-mortem data is residue, not continuity, meaning no current legal or ethical framework gives institutions a clear right to repurpose a person's recorded words and writings for interactive AI.¹ Without a documented posthumous directive or estate approval, the tool operates in a consent vacuum.

Fidelity: The Ask Eddie AI carries a disclaimer that its output is not historically accurate. That acknowledgment points to a deeper risk: the model may misrepresent Bernays' views, smoothing over nuances or fabricating responses that feel authentic but deviate from his actual philosophy. For scholars and practitioners, using such outputs uncritically can erode the integrity of PR history.

Instrumentalization: Deploying Bernays' persona as a museum attraction or educational novelty raises concerns about instrumentalizing a legacy for branding.⁴ Unlike deepfakes of living figures, which often cause immediate reputational harm, historical simulations risk turning a person into a perpetual commercial asset without any ongoing consent. The line between honoring a pioneer's ideas and exploiting his identity can blur quickly.

Guardrails for Responsible AI Persona Tools

Tools like Ask Eddie can adopt concrete guardrails: explicit labeling of AI-generated content, disclosure of data provenance, and clear separation from authentic historical record.¹ IEEE and ACM frameworks further suggest that designers should assess whether the simulation respects human dignity and avoids emotional harm, risks already documented in studies of griefbots and digital necromancy.⁵ A responsible approach would also involve the Bernays estate in an ongoing consent conversation and publicly limit the tool's use to reflective, educational, non-commercial contexts.

Is PR Being Replaced by AI? What Bernays' Ideas Suggest

AI is not replacing public relations. The data shows it is reshaping the toolkit, but the strategic and relational core of the profession remains stubbornly human. In 2025, 77% of PR professionals reported using AI tools in their work, and 59% said AI's importance will grow significantly over the next two years.¹ Yet the same surveys reveal a clear line: AI accelerates execution, not judgment.

The Data on AI Adoption in PR

• Adoption: 77% of practitioners already use AI daily for tasks like media monitoring, drafting press materials, and analyzing sentiment.¹
• Efficiency gains: Agencies report campaign execution times slashed by 70%, manual research hours reduced 60 to 80%, and media list accuracy improving 73%.²
• Early warning: AI-driven monitoring now detects reputational threats 12 to 48 hours earlier than traditional methods.²

These numbers are striking, but they concentrate in operational zones. Where AI shines is in speed, pattern recognition, and scale: drafting 10 pitch variations, scanning millions of articles, or spotting a crisis before it trends.

What AI Excels At (and What It Doesn't)

Tasks being automated at scale include:

• Media monitoring and clipping analysis: AI scans and categorizes coverage 3 to 5x faster.²
• Routine content creation: Press releases, social copy, and email pitches can be drafted in seconds.
• Predictive analytics: Machine learning models forecast story virality or stakeholder reactions.

But AI still stumbles where context, ethics, and relationships matter. Media relations remains 84% of a PR professional's core function, and that function depends on:¹

• Knowing which journalist needs an exclusive versus a background briefing.
• Reading the room during a live crisis to decide whether silence or speech is the right move.
• Building trust over years, not optimizing a click-through rate.

Public trust in AI itself is fragile: only 41% of people say they trust responsible AI use.³ So a brand that delegates its voice entirely to algorithms risks alienating the very publics it hopes to persuade.

Bernays' Vision of PR as a Human-Centered Craft

Edward Bernays famously defined public relations as a "two-way street" between institution and public. For him, the craft relied on deeply understanding human psychology, not just demographics or sentiment scores. His concept of the "engineering of consent" required nuanced judgment about societal values, cultural shifts, and ethical boundaries. AI can surface patterns from vast datasets, but it cannot replicate the moral reasoning Bernays demanded when he said, "Never to accept a position that puts duty to a client above duty to society."

That ethical core is exactly where automation falls short. Crisis communication, stakeholder negotiation, and strategic reputation building all require a human ability to weigh competing interests, interpret ambiguous signals, and make decisions that algorithms would treat as optimization problems. Professionals looking to sharpen those judgment calls will benefit from understanding how PR, marketing, and strategic communication differ at a foundational level.

The Real Career Risk: Falling Behind the Toolkit

For practitioners, the danger is not that AI will replace PR. It is that professionals who ignore AI tools will find themselves outpaced by those who use them. Just as Bernays' social-science approach displaced press-agentry puffery, today's data-literate communicators are replacing guesswork with evidence. The 77% usage figure suggests that ship has already sailed. Ignoring AI now is like refusing to use email in 1998.

But embracing AI does not mean ceding strategy to a machine. It means delegating the repetitive so you can focus on the relational. Those exploring online master's in media communication programs will find curricula increasingly designed around this balance. The professionals who thrive will be those who combine Bernays' humanistic insight with AI's analytical power, using the tool to understand publics more intimately, not just to talk at them faster.

AI Persona Tools Compared: Ask Eddie in Context

The field of AI-generated historical personas is evolving rapidly, moving from novelty entertainment to tools with serious pedagogical potential. As more platforms allow users to chat with simulated figures from the past, understanding how each tool handles training data, accuracy, and audience becomes essential for educators and practitioners.

The Edutainment Divide: Consumer Apps vs. Curated Tools

At the lighter end of the spectrum sit broad-consumer apps like Hello History and Character.AI. Hello History offers a mobile app with a menu of historical figures and free-form chat, built on generic large language model impersonation without a dedicated figure corpus.¹ Its limited guardrails have led to serious distortions, including reported instances of Holocaust denialism in simulated conversations, making it unsuitable for classroom use.² Character.AI similarly allows users to create historical characters through prompt engineering, with no systematic grounding in primary sources and no built-in provenance features; safety and accuracy concerns further restrict its educational viability.² These tools prioritize conversational fluidity over factual reliability, positioning them squarely as informal edutainment.

In contrast, platforms designed with scholarly or instructional rigor take a fundamentally different approach. Humy, built for K-12 social studies, employs retrieval-augmented generation over curated primary and secondary sources, with teachers able to extend the corpora.³ It emphasizes provenance, includes controls for sensitive topics, and is aligned with standards-based instruction. Museum-led projects, such as chatbots simulating Nikola Tesla or Branislav Nušić, rely on verified corpora drawn from letters, autobiographies, archival interviews, and scholarly works.² These simulations explicitly label their outputs as constructed, with strong transparency about limits and expert oversight. Their audience is museum visitors and students seeking culturally authentic interactions, not casual chat.

Where Ask Eddie Fits: A Museum-Backed, Profession-Specific Model

Ask Eddie occupies a distinctive middle ground with a clear professional tilt. Developed by The Museum of Public Relations, it draws on Edward Bernays' recorded interviews and published writings, a deliberately curated training set. This mirrors the museum-chatbot approach of grounding responses in a defined, transparent corpus. The tool carries an explicit disclaimer that outputs are not historically accurate for academic citation, a level of disclosure that consumer apps rarely match. Its intended audience is PR practitioners, scholars, and students, not the general public, which shapes both the depth of content and the ethical framing. For professionals weighing how these tools intersect with broader strategic communication case study work, that specificity matters.

Interactivity and Educational Rigor Across Tools

Regarding features, Ask Eddie and most museum chatbots offer direct, open-ended conversation with the historical figure, while Humy layers on teacher assignment generators and safety controls.³ Ask Eddie lacks the classroom governance features of Humy, but its narrower, expert-driven corpus may yield more nuanced answers on public relations topics than a general edutainment bot. The free-form interactivity of Ask Eddie aligns with professional reflection tools, where users pose complex ethical or strategic questions rather than following a structured lesson. Practitioners interested in staying current with emerging tools like these should explore latest trends in communication across the field.

What Makes Ask Eddie Unique in the PR Landscape

The defining differentiator is topical focus. No other historical persona AI is purpose-built for the communication field. While generic tools might attempt to answer PR questions by extrapolating from broad internet text, Ask Eddie's responses are technically constrained, and enriched, by Bernays' actual words and ideas. This positions it as a reflective instrument for exploring the "engineering of consent" and two-way communication principles in their original context, with an embedded caution that even these simulations require critical evaluation.

Using AI Bernays as a Teaching and Reflective Tool

Using the Ask Eddie AI in educational settings means treating the simulated Edward Bernays as a conversation partner who sparks critical thinking rather than a definitive historical source. The tool works best when it challenges assumptions and generates questions, not when it provides settled answers.

Four Ways to Integrate Ask Eddie into Learning Environments

• Graduate Seminar Exercise: Students pose questions about core concepts like "engineering of consent," then compare the AI's responses to primary-source readings from Bernays' books and recorded interviews. Discrepancies often highlight nuances in his evolving thought.
• Strategic Planning Warm-Up: Before tackling a modern campaign dilemma, a team "consults" Bernays by asking how he might approach a similar challenge in his era. The AI's answers surface timeless principles, which the team then adapts to today's media landscape.
• Professional Development Reflection: Practitioners reflect on ethical foundations by posing current challenges to the AI. Contrasting Bernays' mid-20th-century views with modern complexities illuminates both enduring values and new tensions.
• Ethics Classroom Exercise: Students query the AI on ethically charged topics, then cross-check responses against documented historical records and speeches. They track alignments and divergences, building research and media literacy skills.

The Transcript as a Teaching Artifact

The tool's downloadable transcript feature is particularly useful in the classroom. Students can annotate each AI-generated claim, comparing it to source material from Bernays' actual writings and interviews. Submitting the annotated transcript as an assignment directly reinforces the habit of verifying AI outputs against primary evidence. For graduate students preparing applications, the same careful sourcing discipline applies when crafting a communication graduate school statement of purpose.

Heeding the Citation Warning

Both the tool and its creators explicitly warn that Ask Eddie's outputs are not historically accurate sources for academic purposes.¹ In any teaching context, instructors must emphasize that citing AI Bernays as a historical authority violates academic integrity. Instead, the tool serves as a prompt generator that drives learners back to original research, turning a potential shortcut into a deeper engagement with the primary record. This kind of disciplined evaluation is a skill that strengthens successful marketing communication strategies as well as academic work.

Common Questions About AI Bernays and the Ask Eddie Tool

Will AI-simulated interviews become a standard tool in PR education and professional development? Ask Eddie shows the potential but also the limits. The tool opens an interactive window into Bernays' thinking, yet the Museum of Public Relations' own disclaimer reminds users that its answers are not a historically accurate source. That tension is the real lesson: AI personas can spark reflection, but they demand critical judgment. Try the tool, compare its responses to Bernays' actual writings, and join the profession's growing conversation about how to ethically govern AI representations of its own history. For professionals committed to communicating effectively in the workplace, these AI-powered reflections on PR's origins offer a sharper lens on timeless principles. The next chapter of PR's past is being written with AI, but practitioners must hold the pen.