Deepfake Democracy

I’ve worked in banking for 23 years, with nearly half of that time focused on fraud risk. Over the years, I’ve seen fraud evolve in both scale and sophistication, but one moment marked a turning point. I came across a newspaper headline that read: “Deepfakes are coming, and we (banks) aren’t ready.” This struck a nerve. This wasn’t just about new technology, it was about the fundamental question of whether we, as institutions built on trust, were prepared for what was coming. That moment sparked my PhD journey, which explores the implications of deepfakes on public trust, in society more broadly, but especially within banking.

This research sits at the intersection of disciplines, supervised by both a social scientist and a computer scientist. That balance has been essential. Early in the process, I made what I thought was an uncontroversial, unequivocal even, observation: “Fraud has never been as bad as it is today.” My social science supervisor immediately pushed back. Is it worse, or is it a matter of how we define fraud? How we measure it? How we choose to report it?

That challenge fundamentally reshaped my approach. To understand whether fraud today is truly exceptional, or simply differently framed, I began digging into the Hansard database of parliamentary records, which document political debate in the UK since the early 1800s. I wanted to understand how fraud has been talked about over time: when it’s seen as a crisis, who is blamed, and how those moments shape public perception, policy, priorities and reactions.

What I found was that “Fraud”, something we often associate with modern cybercrime or regulatory lapses, can be deemed to have gone through three major conceptual phases in the parliamentary record. Each phase tells us not only how elected officials have responded to fraud, but what society has expected of them, and what it has often overlooked.

1) The Moral Phase: Fraud as a Character Defect

In the 19th and early 20th centuries, fraud was framed in starkly moralistic terms. Parliamentary debate presented it as a personal failing, a betrayal of one’s social or professional role. Fraudsters were spoken of not only as criminals but as dishonourable, immoral figures who had violated the implicit trust of society. These were debates heavy with Victorian notions of virtue and shame. Fraud was seen as corrosive not just because of monetary loss, but because it offended the moral fabric. In these earlier debates, fraud was intimate, it happened in the parish, the marketplace, the stock exchange. It was a human crime, and its remedy was as much reputational as judicial.

The March 1809 debate in the House of Commons concerning the “Conduct of the Duke of York” was an example of how accusations of fraud were treated at that time as not merely legal offenses but as matters deeply rooted in morality and public trust. Prince Frederick, Duke of York and Albany, then Commander-in-Chief of the British Army and the second son of King George III, was implicated in a scheme involving his former mistress, Mary Anne Clarke. Clarke alleged that she had sold military commissions in exchange for money and gifts, and that the duke was aware of, and perhaps complicit in, these transactions. Although the inquiry ultimately found no conclusive evidence that the duke personally benefited from the corruption, the suggestion that he allowed, or ignored, such misconduct severely damaged public confidence in both the military and the monarchy. The resulting scandal was not treated as a narrow legal issue but as a broader moral failure, highlighting how the concept of fraud has long functioned as a reflection of societal expectations around integrity, power, and institutional responsibility.

The duke resigned from his post in March 1809 but was later reinstated in 1811 after the inquiry concluded. The episode illustrates how debates around fraud have historically been shaped by concerns about reputational damage, elite accountability, and the moral fabric of public life, themes that continue to resonate in contemporary discussions. One speaker noted that the evidence against the duke, was largely undisputed in terms of fact, but the critical issue was whether the duke was aware of the wrongdoing. The matter went beyond legality, and became one of a breach of trust, a betrayal of the moral duties owed by someone in high office. Also, during that debate, William Windham, exclaimed that “The moral part of the question, as it is called, is one… that many gentlemen think ought of itself to call for the animadversion of the house” pointing to how moral culpability, or absence thereof, is a distinct and vital consideration. It's not sufficient to ask, “Was a rule broken?” The key question is: “Did they betray moral standards expected of their rank”. Speakers talked also of public trust and the honour associated with royal office. The discourse points to two intertwined ideas of how firstly the House must strive both legally and morally to uphold public confidence and that even the appearance of fraud, or connivance in it, was sufficient to undermine an institution, regardless of criminal guilt.

The Adulteration of Food and Drink Bill debate in 1860 was another such example where the fraud was treated not simply as a technical issue surrounding safety, but a deeply moral one. The core concern was that food and drink were being deliberately tampered with in ways that endangered health and exploited the public, particularly the poorer classes. Substances such as lead, copper, and even arsenic were added to enhance the appearance or shelf life of products. Lead in wine and cider, copper compounds to green vegetables, while staples like milk were routinely diluted with water and thickened with chalk or starch. Bread was whitened using alum or other chemicals, and everyday items like tea, spices, and mustard were often bulked out with fillers ranging from flour to brick dust or even ground nutshells. These adulterations not only degraded nutritional value but often introduced harmful toxins, all while masquerading as legitimate goods.

What made the issue particularly concerning was that it disproportionately affected the working classes, who lacked both the resources and the scientific knowledge to protect themselves. As Mr. Wise noted in the House of Commons on 29 February 1860, “systematic adulteration… endangered public health, practised pecuniary frauds extensively, and deeply affected the moral character of the country.” The practice was thus framed not merely as a commercial or safety issue, but as a profound moral failing, a betrayal of the most vulnerable in society, and a stain on national integrity. (Mr.Wise 29 Feb 1860). 

2) The Regulatory Phase: Fraud as a Systems Failure

By the mid-to-late 20th century, the tone changed from one of moral indignation to rather framing fraud as a systemic failure. In the wake of scandals like the secondary banking crisis, the collapse of Barings, and the Maxwell pension thefts, fraud was reframed as a systems problem. It became a matter of institutional weakness, regulatory oversight, and statutory enforcement.

Parliamentary debates shifted from moral diatribe to more technical language: due diligence, compliance regimes, regulatory capture, statutory instruments. The rise of the Financial Services Authority (and later the FCA and PRA) marked a clear delegation of responsibility from Parliament to arm’s-length bodies.

In these moments, fraud lost some of its personal drama, it became a governance challenge, something to be "managed" through frameworks, risk registers, and disclosures. While this shift professionalised the response, it also distanced political actors from direct responsibility. Fraud became a problem for the regulators, not a societal condition to confront.

3) The Technological Phase: Fraud as an Arms Race

In the last decade, a third framing has emerged, one shaped by our increasingly digital lives. Fraud is now often depicted as a technological arms race, dominated by concerns around the scale and anonymity that the internet has introduced and the fear of the hacker or identity thief.

The debate is more urgent, more reactive, and more visibly unsettled. “They can operate anywhere in the world, know everything about their chosen victim and be seemingly invisible to detection. This AI-assisted crime is growing and becoming ever more sophisticated”. (Dean Russell, 22 Jan 2024). During a debate about revenge porn Maria Millar declares “the days of treating the internet as the wild west are, I am glad to say, long gone.” (Maria Millar, 19 Jun 2014) This was reflective of a growing parliamentary consensus that the previously unregulated internet space was becoming untenable and needed to be subject to enforcement and oversight, particularly in areas like online abuse and privacy. The language being used became more emotive AI infused scams resulting in “tragedy”.

Where fraud was once about the dishonest trader or the complicit banker, it' is now about algorithmic amplification, faceless threat actors, globalised deception networks and the associated metaphors shifting from the courtroom to the command centre.

What Hasn’t Changed?

Despite this evolution, one uncomfortable pattern remains unwavering. Fraud rarely receives serious political attention until after public outrage peaks. Whether it’s the appearance of “synthetic” butter in food stores, collapse of a household-name institution, a surge in elderly victims of romance scams, or tabloid headlines about synthetic voices draining pension pots, it is reaction, not prevention, that drives parliamentary urgency.

Even in 2025, with record-high scam losses and growing concern over trust erosion, prevention is often discussed only after the crisis has hit the front page. This latency matters, because today’s fraud is not simply a financial crime. It is an attack on societal infrastructure. It erodes trust in financial institutions, in digital interactions, and trust in people and reality itself. When a deepfake video can fool an accountant, or a chatbot can socially engineer a bank customer into transferring their life savings, the question isn’t just "how was that fraud able to happen?" but "what else is now breakable?"

The New Cost of Fraud: Trust

My PhD work focuses on how deepfakes, synthetic media that simulate people’s likenesses and voices, are reshaping public trust in banking. But this technology is only the tip of the spear. I think what we're witnessing is the scaling of deception, where traditional safeguards like visual confirmation, voice recognition, or even intuition no longer provide protection.

In this context, the consequences of fraud aren’t just financial. They are epistemic. When people can no longer trust their eyes, ears, or institutions, trust becomes the primary casualty. And with that, the foundations of digital society begin to erode.

Where Do We Go from Here?

The question we must ask ourselves is whether Parliament is speaking the right language about fraud?

• If fraud is now a problem of information disorder, should responses be led solely by financial regulators?

• If deception is now aided and abetted by the infrastructure of the internet, should platforms be held to different standards?

• If victims are now systemically targeted, should public protection be proactive, not compensatory?

Fraud is no longer a marginal issue. It is a core governance challenge, a public safety issue, and a societal risk. The Hansard corpus reflects this growing tension, but also reveals the persistent lag between recognition and response. 

Let’s Shift the Frame

It’s time to move beyond reactive cycles of scandal and outrage. We need a fraud policy conversation that is:

• Forward-looking, recognising that emerging technologies will outpace current laws and controls.

• Holistic, bridging regulatory silos between financial services, digital platforms, and national security.

• Moral, in the truest sense and focused on what kind of digital society we want to build and protect. 

If we don’t shift the conversation now, we may find ourselves debating the next wave of fraud only after the public’s trust is already lost and countless people have been harmed.

Policymakers, technologists, banks, regulators, and the public all must work together else we are destined to always be one step behind fraud, rather than finally leading from the front.

MJ is a career banker turned fraud strategist who treats financial crime not just as a threat vector but as a cultural signal. Having spent two decades designing controls, decoding scams, and defending the financial perimeter, she now splits her time between regulatory war rooms, scam prevention campaigns, and a PhD on synthetic media and societal trust. For her, fraud isn’t just a matter of money, it’s a system failure in human judgment, technological design, and institutional memory. She believes good fraud defence requires more than tech stacks and trigger words. It needs empathy, anthropology, and a sharp eye for how people really behave under pressure. Drawn to clarity in chaos and allergic to jargon, she sees fraud as less about bad actors, and more about blind spots in how we design for human weakness.