WEB DESK: Sweden’s government is pushing forward with a new proposal that could significantly enhance the police’s ability to track criminals using AI-powered facial recognition technology. The bill would allow law enforcement to deploy cameras capable of identifying wanted individuals in real-time, even in large crowds, in public places such as streets, squares, and train stations.
The Ministry of Justice has submitted the legislative proposal for consultation today, which aims to empower police with AI systems for automatic facial recognition. Under this system, if a suspected criminal’s image is available, it can be entered into the AI database, allowing surveillance cameras to scan public spaces for that individual, regardless of whether they are already known to authorities.
Justice Minister Gunnar Strömmer (M) emphasized the potential for this technology to significantly enhance the police’s ability to intervene in situations involving serious crimes like shootings or explosions. “By receiving an immediate alert when such individuals pass a camera, the police can act swiftly to prevent further harm,” Strömmer stated.
The proposal has been fast-tracked by the government in response to an EU regulation that took effect last summer, which allows for more lenient use of such technologies. The police have expressed strong support for the bill, citing its potential to greatly enhance crime detection and prevention.
However, the proposal has sparked concerns regarding privacy. Critics warn that the widespread use of AI-powered facial recognition could lead to significant invasions of personal privacy, as it involves the surveillance of large numbers of people in public spaces. “Such systems could represent a major intervention into personal privacy, enabling constant monitoring of the public,” the government’s investigator noted in a report last year.
In response to these concerns, Strömmer reassured that the new system would be subject to strict legal safeguards, ensuring its use is limited and regulated.
Under the proposed legislation, the use of AI-powered facial recognition will be restricted to identifying individuals suspected of committing serious crimes, such as murder, rape, or major weapons offenses, which carry sentences of at least four years in prison. It will also be used to help prevent crimes against human life, such as terrorist attacks. For example, the system could help identify known terrorists entering the country or track down missing persons who may be victims of human trafficking or kidnapping.
The technology will only be used to locate and identify specific individuals, and not for general crime prevention or broad surveillance. Any real-time facial recognition action will require approval from a prosecutor or court, adding an additional layer of oversight.
To implement this new technology, the police will need to develop new technical solutions and replace existing cameras that are not AI-compatible. According to the police, they anticipate needing approximately 100 permits for AI camera usage in the first year, with the number expected to rise to around 500 per year within a few years, potentially growing to 1,000 annually as the system becomes more widespread.
As Sweden moves forward with this groundbreaking legislation, the debate over the balance between enhancing public safety and protecting privacy is set to continue.
