UK Secretly Researching Murderer Prediction System, Is the Plot of “Psycho-Pass” Coming True?
In the Japanese anime “PSYCHO-PASS,” there is a powerful monitoring network called the “Sybil System,” which actively scans citizens’ mental and psychological states, calculating what is known as the “Mental Index” and “Crime Index,” predicting an individual’s likelihood of committing a crime. If the index exceeds a certain threshold, even if the person has not yet committed a crime, they may be pursued by the police or even executed on the spot.
The Japanese anime “PSYCHO-PASS” features a Sybil System that calculates the crime index of citizens.
Dystopian sci-fi settings like this are often used in fictional media, such as “Black Mirror,” “Minority Report,” and the aforementioned “Psycho-Pass.” However, with the continuous advancement of technology, the future of relying on AI for social safety management may indeed be approaching.
A recent revelation by the non-profit research organization Statewatch shows that the UK Home Office is secretly conducting a “Homicide Prediction Project,” attempting to predict who may commit murder in the future by analyzing large amounts of personal data.
What is the UK Homicide Prediction Project? How Does AI Determine Murderous Tendencies?
Statewatch obtained a government document through the Freedom of Information Act, revealing that the UK is promoting the “Homicide Prediction Project,” commissioned by the Prime Minister’s Office and in collaboration with the UK Home Office, the Greater Manchester Police, and the Metropolitan Police Service. This homicide prediction system utilizes data from the UK Home Office, the National Police Computer system, and the Greater Manchester Police for what is termed “homicide risk assessment.” The Greater Manchester Police have provided personal data for between 100,000 and 500,000 individuals.
The UK secretly researches a murderer prediction system, sharing personal data of 500,000 people between police departments and the Home Office.
The UK Home Office states that health marker data is expected to have significant predictive power, encompassing sensitive information related to people’s mental health, addiction, self-harm, suicide, vulnerability, and disability. Research documents reveal that the Home Office’s data science team will “develop models” to identify strong predictors of homicide risk within the data, suggesting that this crime prediction system may begin operating in the future.
Is It Really Okay to Use AI to Predict Murderers?
Using crime prediction systems to determine whether someone will commit a crime is a highly controversial moral endeavor. Statewatch researcher Sofia Lyall has expressed that this is “chilling” and “dystopian.” However, what is even more concerning is that crime prediction systems built with AI algorithms may classify individuals who have not yet committed crimes as “high-risk,” subjecting them to unnecessary surveillance. “Research has repeatedly demonstrated that algorithms used for predicting crime are inherently flawed. Yet, the government is vigorously promoting AI systems that can label individuals as criminals before they have committed any act,” she stated.
Research conducted by the Home Office itself found that the recidivism prediction tool “Offender Assessment System” (OASys) had a lower predictive accuracy for Black offenders compared to White offenders, yet this system is currently used for sentencing decisions, prison classification, and parole decisions. Lyall criticized that utilizing models based on UK police and Home Office data, which are inherently racially biased, would reinforce and amplify structural discrimination within the criminal justice system. Furthermore, using sensitive data related to mental health, addiction, and disability is an extreme invasion of people’s privacy.
Is an AI Surveillance Society Imminent? Researchers Urge: Halt Homicide Prediction Research Immediately!
Currently, while the UK’s secret research into crime prediction systems is still in the research phase, the government must contemplate how to balance crime prevention with the protection of citizens’ privacy and freedoms as AI technology evolves. From an ethical standpoint, labeling someone as a “potential murderer” can create indelible social stigmas, especially since the current predictive technology still exhibits biases and inaccuracies.
When crime becomes difficult to prevent, and the government, while cutting social welfare, still harbors the delusion that technology can solve everything, it may encounter issues of superficial solutions without addressing the root causes. Statewatch researchers call for:
“The UK Home Office must immediately cease further development of homicide prediction tools. The government should not spend money developing problematic and racially discriminatory AI and algorithms, but should invest in genuinely supportive social welfare services.”
If a future with AI surveillance indeed comes to pass, then the plot of “PSYCHO-PASS” may become a chilling prophecy. As mainstream media outlets like The Guardian report on this research, public discussions are already underway, prompting people to ponder: “As the plots of science fiction and anime gradually become reality, is our society prepared?” And in a society under long-term systemic control, how will people’s freedoms and choices be preserved?
This article is authorized for reprint from: Encrypted City