Artificial Intelligence and Synthetic Reality
When a familiar voice can be imitated and a convincing face can be manufactured, the appearance of evidence is no longer sufficient proof.
The Biblical Foundation
Exodus 20:16 - "Thou shalt not bear false witness against thy neighbour."
The commandment establishes a moral responsibility concerning truthful testimony. Modern technology has dramatically expanded the means through which false representations can be created and distributed. Artificial intelligence can generate convincing images, voices and videos, including material that appears to document events that never occurred. The technology is new, but the underlying moral questions concerning honesty, impersonation and false accusation are not.
The Question Before Us
What happens when seeing a person's face or hearing their voice no longer establishes that they said or did what a recording appears to show?
Artificial intelligence can produce increasingly realistic synthetic media. These capabilities support legitimate work in education, accessibility, research and creative industries. They also enable fraud, impersonation, fabricated evidence and political misinformation. This investigation examines how synthetic media is created, why it is persuasive and what individuals and institutions can do to verify digital evidence.
A New Problem for an Old Assumption
For much of the history of photography and recorded sound, people have treated images and recordings as important forms of evidence. They were never completely reliable.
- Photographs could be staged or altered.
- Audio recordings could be edited.
- Videos could omit important context.
But producing highly convincing fabricated material generally required specialised skills and resources. Generative artificial intelligence has substantially reduced some of those barriers. A person can now produce realistic images, imitate voices and generate video content using increasingly accessible tools.
This changes the practical difficulty of verifying what appears on a screen.
How Synthetic Media Creates Credibility
A convincing deepfake may combine several elements.
- A familiar face creates recognition.
- A copied voice creates apparent authenticity.
- A realistic setting makes the recording seem plausible.
- A caption supplies the intended interpretation.
- The viewer may respond emotionally before considering whether the material is genuine.
This is particularly dangerous when the content appears to show a respected person making an inflammatory statement, confessing misconduct or endorsing a financial opportunity. The more familiar the person appears, the more persuasive the impersonation may become.
Artificial Intelligence and Financial Fraud
Voice cloning can be used to impersonate family members, executives or public officials. A fraudulent caller may claim that a relative has been arrested or urgently needs financial assistance. A business may receive an apparent instruction from a senior executive to transfer money. The familiar voice becomes part of the deception.
This means that recognising someone's voice is no longer sufficient verification for a high-risk financial request. Independent confirmation through a previously established communication channel becomes increasingly important.
Artificial Intelligence and Religious Claims
Religious communities may be particularly vulnerable to fabricated material involving respected spiritual figures. A synthetic video could appear to show a preacher endorsing a doctrine he has never taught. A fabricated recording could appear to document a miracle. An impersonated religious leader could request donations or claim to possess urgent prophetic information.
These possibilities do not mean that every unusual religious recording is false. They mean that visual and audio evidence must be examined in context. Original files, independent witnesses, publication history and reliable corroboration become increasingly important.
Deception in Practice - Documented Misuse of Synthetic Media
Case Study 1 - AI-Generated Biden Robocalls in New Hampshire
In January 2024, voters in New Hampshire received robocalls using an AI-generated imitation of President Joe Biden's voice. The calls discouraged participation in the state's presidential primary. The US Federal Communications Commission subsequently imposed a $6 million penalty against the political consultant responsible for the illegal calls. The case demonstrates how a recognisable voice can be used to create false political authority.
The recipient did not need to believe an elaborate conspiracy. The deception depended upon the apparent authenticity of a familiar public figure.
Case Study 2 - The Hong Kong Deepfake Video Conference Fraud
In 2024, Hong Kong police reported a case in which a company employee was deceived through a video conference involving synthetic representations of colleagues. The incident resulted in transfers totalling approximately HK$200 million. The case demonstrates that synthetic impersonation can operate through several apparent participants rather than a single copied voice.
It also illustrates how fraudsters can exploit normal workplace procedures and expectations. The apparent presence of familiar colleagues can reduce the likelihood that an employee will question an unusual request.
The Opposite Danger - Dismissing Genuine Evidence as Fake
Synthetic media creates another problem. As people become aware that recordings can be fabricated, genuine evidence may be dismissed as artificial. A public figure confronted with an authentic recording might claim that it was generated by AI. A person accused of wrongdoing may argue that photographs or audio have been manipulated.
This possibility is sometimes called the liar's dividend. The existence of deepfake technology does not mean that every disputed recording is false. It means that authentication methods become more important.
What Responsible Verification Looks Like
Verification should begin with the source.
- Who originally published the recording?
- Is the original file available?
- Can the event be confirmed through independent witnesses?
- Does the material contain inconsistencies?
- Are there trustworthy records showing that the person was present?
Technical detection tools can assist, but they are not infallible. A detector's output should not be treated as conclusive proof without considering its reliability and the wider evidence. For financial requests, the safest procedure may be to contact the supposed sender independently using a known telephone number or established internal process.
Beyond the Church - Technology Is Not the Moral Agent
Artificial intelligence can assist with medical research, translation, accessibility and education. It can also be used to produce misleading information. The technology does not possess moral accountability in the same sense as the people and institutions developing, deploying and using it.
The ethical questions concern human decisions.
- Who created the material?
- Was it intended to mislead?
- Was consent obtained where appropriate?
- Could the output cause foreseeable harm?
- How should responsibility be assigned?
Christian ethics approaches these questions through principles of truthful witness, justice and responsibility towards others. Non-Christian ethical frameworks may reach overlapping conclusions through privacy, autonomy, harm prevention and human rights.
What Have We Established? - Verification Must Replace Assumption
Synthetic media has weakened the traditional assumption that a convincing recording is reliable evidence of an event. The risks extend across politics, finance, religion, personal relationships and criminal investigations. But the appropriate response is not to distrust every image or recording automatically. It is to improve verification.
A convincing appearance is evidence to examine, not a conclusion to accept. And an allegation that something is AI-generated must itself be supported. In a world where falsehood can look and sound authentic, truth requires more than trusting our eyes and ears.
Further Study
Further Scripture
Exodus 20:16 | Proverbs 12:22 | Proverbs 18:13 | Matthew 24:4-5 | Ephesians 4:25 | Revelation 12:9
Related investigations
Self-Deception | Our Commitment to Truth | Political Manipulation | Media Manipulation