How to spot a deepfake—and prevent it from causing political chaos
The “fake news” of 2016 is so passé. This year, fake video and audio clips generated by artificial intelligence (AI) are the new looming threat that could sway voters in the upcoming U.S. presidential election. Sourse: Science.org
A robocall that used an AI voice resembling President Joe Biden’s to advise New Hampshire voters against voting in the state’s presidential primary has been linked to a pair of Texas-based telecommunications companies, the state’s attorney general announced on Tuesday. Source: CNN
Objectives:
- To raise students’ awareness of deepfakes and their potential risks to democracy and society.
- To develop students’ critical thinking and media literacy skills by analyzing and evaluating different sources of information.
- To practice students’ speaking and writing skills by expressing their opinions and arguments.
Tasks
- Read the text below,
- Watch the video
- Look at the pictures
- Answer the questions below
- In groups, discuss this topic using the discussion prompts below
- Try making a deep fake yourself and test it in your class
- Choose one of the essay questions below.
- What is a deepfake, and how is it different from other forms of digital manipulation?
- What are some of the reasons why people create and share deepfakes?
- What are some of the challenges and dangers of detecting and regulating deepfakes?
- What possible solutions or strategies exist to combat deepfakes and protect the public from misinformation?
Discussion Prompts:
- Do you think deepfakes are a severe threat to democracy and society? Why or why not?
- How do you evaluate the credibility and reliability of online information? What are some of the signs or clues that indicate a source is trustworthy or not?
- How do you react when seeing or hearing something contradicting your beliefs or opinions? Do you verify the facts, or do you ignore or dismiss them? Why?
- How can you educate yourself and others about deepfakes and media literacy? What are some of the skills or habits that you need to develop or improve?
Essay questions:
- How do deepfakes pose a threat to democracy and society? Provide examples of how deepfakes have been used to spread misinformation and disrupt public discourse. Evaluate the effectiveness of the current solutions or strategies to combat deepfakes and protect the public from misinformation.
- Define media literacy and why it is essential in the digital age. Discuss how you evaluate the credibility and reliability of the information you encounter online. Explain how you can educate yourself and others about deepfakes and media literacy.
- How do deepfakes affect the trust and credibility of the media and public figures? Provide examples of how deepfakes have been used to damage the reputation or influence the perception of specific individuals or groups. Discuss the ethical and legal implications of creating and sharing deepfakes without consent or disclosure.
Is deepfake illegal?
AI-generated deepfakes provide a “perfect storm” for malicious states and criminals who want to hijack the general election, the home secretary has warned.
Speaking ahead of meetings with Silicon Valley bosses where he will urge greater action to protect democracy, James Cleverly said the rapid advancement of technology combined with the mass use of social media poses a serious threat to elections being held around the world this year.
He is expected to meet the bosses of Google, Meta, Apple, YouTube and other major tech companies in San Francisco to tell them to take collective action to protect democracy.
Cleverly warned that “malign actors” — including those acting on behalf of hostile states such as Russia and Iran — and criminals will attempt to use generative AI to quickly produce thousands of illicit images or deepfakes.
They can then be shared on encrypted services with impunity and shared with millions on social media within seconds, he said. Source: The Times
Is the impact of deepfakes truly more severe than the traditional spread of misinformation? Misinformation on social media platforms has been a longstanding issue. Studies have shown that individuals are prone to disseminating false news articles that align with their own viewpoints, even when they are aware of the falsehoods. The labeling of content as fake news does not deter this behavior: the more we are exposed to false content, the more likely we are to recall it as true. “Repeated exposure embeds content in people’s minds,” states Steven Sloman, a cognitive psychologist at Brown University.
However, this issue is exacerbated with the use of images and videos, which tend to leave a more lasting impression than text. “We are naturally inclined to believe what we see,” Sloman explains. Therefore, convincing deepfakes, even if only briefly viewed, could potentially be shared and accepted more widely.
Sloman points out another risk associated with deepfakes: they breed uncertainty. If individuals are unable to distinguish between real and fake images, they may be more inclined to label genuine images that contradict their beliefs as deepfakes, irrespective of expert analysis or detection software. A study revealed that warning people about the potential AI generation of videos did not improve their ability to identify them. On the contrary, it made them more skeptical of all visual content they encountered. In this regard, Sloman asserts, “Deepfakes pose a greater threat than any other medium that has existed before.” Source: Science.org
An estimated 2 billion people are expected to vote in national elections this year in the UK, US, India, and 60 other countries. Deepfake technology has been prevalent, with audio imitating Sir Keir Starmer and Sadiq Khan and videos featuring Rishi Sunak in fabricated news. These deepfakes have been used to spread misinformation and disrupt public discourse.
Tech firms and governments are urged to take collective action to prevent these malign actors from influencing the elections. The digital sphere, where these deepfakes are inserted, needs rules, transparency, and user safeguards.
Technological acceleration has made sophisticated digital tools available to malign actors worldwide. Google has established “24/7 war rooms” to identify misinformation, recognizing the severe threat posed by AI-powered deepfakes to democracy.
A well-timed deepfake could influence a close vote, as seen in the 2020 US presidential race. Households in New Hampshire received robocalls from an audio deepfake voice that sounded like Joe Biden urging people not to vote. A survey found that 70% of MPs fear AI will increase the spread of misinformation and disinformation in the next general election. The UK’s new Online Safety Act does not outlaw deepfakes.
One comment