Ethical, Legal, and Societal Implications

As you settle into your usual routine of scrolling through emails, you stumble upon a video that sends a chill down your spine. In disbelief, you and millions of other users on the internet watch a video of yourself crying and confessing to crimes you never committed. Shocked and confused about what you just witnessed, your phone starts buzzing with notifications from family, friends, colleagues, employers, and officials, all seeking explanations.

Welcome to the era of deepfakes.

Deepfakes are a form of synthetic media generated through AI techniques, primarily deep learning algorithms, to create or alter visual and audio content so convincingly that it appears genuine. As we navigate the ever-evolving digital landscape, we are confronted with an unsettling reality: what we see and hear may not always be what it seems.

In recent months, headlines have been dominated by chilling accounts of victims falling prey to meticulously crafted illusions. Yet, the threat of deepfakes goes beyond financial fraud or celebrity scandal. With the looming spectre of deepfakes over the political landscape, the very fabric of our governing system hangs in the balance.

One critical question emerges: Should creating deepfake content without a person’s consent be illegal?

It is a question that challenges us to confront the ethical dilemmas posed by rapidly advancing technology. We must unravel the complexities of deepfake’s impact on our society and consider the pressing need for decisive action in an age of deception and uncertainty.

Recent incidents involving Taylor Swift demonstrate the continued evolution of deepfakes. The circulation of sexually explicit deepfake images of Swift, reaching millions of views before being removed, highlights the alarming ease with which the manipulated content can be disseminated online.

The Biden administration has condemned the spread of deepfake content, calling on social media companies to take more responsibility. But is mere condemnation sufficient at this stage? Especially when deepfake technology has evolved to the point of reaching a much larger, mainstream audience.

Consider the current situation in Hong Kong, where an employee fell victim to a deepfake video conference call impersonating senior officers. Deceived by false replicas, the employee unwittingly transferred millions of dollars to fraudsters. This incident spotlights the vulnerabilities inherent in remote communication platforms, which are only becoming more prevalent. Instances of deepfake fraud can have far-reaching consequences for businesses, investors, and the economy at large.

Deepfake manipulations risk causing market instability. False information spread through deepfake content could trigger panic selling, resulting in significant losses for investors and destabilising the broader economy. As investors lose confidence in the integrity of financial markets, capital flows may be disrupted, thus potentially hindering economic growth and development.

This incident emphasises the growing threat to political integrity and democratic processes. Synthetic media leads to a distrust in the political process and corrupted electoral outcomes. It’s become far too easy to spread a fake video of a politician engaging in unethical or criminal behaviour. This technology’s capacity for misinformation, propaganda, and polarising narratives poses significant risks to international relations and societal cohesion.

All instances of deepfakes so far have been inherently negative, spread by malicious actors working for their personal gain while wreaking havoc on the rest of society. There are, however, arguments that support deepfakes.

Just like other technological developments, deepfakes were not created with malicious intent. The scope for positive gains from deepfakes is expansive. For example, deepfakes can create realistic digital doubles of criminal suspects using forensic evidence, thus streamlining the investigation process. Prediction models can also be improved upon, emergency scenarios can be simulated, and it can improve accessibility for individuals with disabilities by enabling the creation of synthetic speech, sign language interpretation, and facial expressions.

Ultimately, deepfakes are a tool. A tool with far too much power to be left in the hands of any individual who utilises it for their own gain. Policymakers and regulators face the daunting task of addressing the unique threats posed by synthetic media. Should this technology be banned altogether? Should we consider the potential benefits that can come from the regulated use of deepfakes? New legislations are in the works, such as the Digital Imprints Regime which requires AI-generated content to have an imprint when uploaded. This is a step in the right direction, but will it be enough? It may prevent misinformation but what about violation of privacy rights or defamation?

Perhaps focusing on regulation by outlawing misinformed deepfakes is a strong first step. Nonetheless, the impacts of deepfakes are intertwined and intricate. Only by working together, across sectors and borders, can we confront the challenges and opportunities posed by deepfake manipulation while ensuring that our rights and values endure for generations to come.

© Lawrence Power 2024