Data protection ombudsman raps Finnish police over controversial facial ID app

Clearview AI uses publicly-posted social media photos scraped from the web for use in law enforcement agency investigations.

Last spring, it was revealed that the NBI had used Clearview AI, even though its use was not widely known within the organisation. Image: Yle

The office of the Deputy Data Protection Ombudsman has issued a reprimand to the Police Board over its use of the controversial Clearview AI facial recognition software.

The programme relies on illegal processing of people's personal data including publicly-posted social media photos scraped from the web in order to help law enforcement agencies in their investigations.

Last spring, it was revealed that the the National Bureau of Investigation's (NBI) had used Clearview AI, even though its use was not widely known within the organisation. Among other applications, the programme can be utilised to enter the images of sexual abuse victims in efforts to identify and protect them.

The NBI has said that the agency's child exploitation investigations unit used Clearview AI in late 2019 and early 2020. During its use, four individuals at the NBI carried out a total of 120 searches on the system over the period of one month.

Deputy Ombudsman: App doesn't comply with Finnish law

The NBI decided to try Clearview AI after it was recommended at a meeting hosted by European law enforcement group Europol in 2019.

In the reprimand issued to the NBI on Tuesday, the Deputy Data Protection Ombudsman noted that the police had used a facial recognition programme without adequately ensuring in advance that the app complied with data security or data protection legislation.

The Finnish Police Board has taken the data protection ombudsman office's reprimand seriously, according to the Board's information management chief Annina Hautala.

"We will also ensure that the measures set out in the [office's] remark are implemented. We are in the process of developing our expertise and procedures so that this does not happen again in the future," Hautala told Yle by telephone.

The office of the Deputy Data Protection Ombudsman instructed the police to notify the individuals in the images that were used in Clearview, if their identities were known. Police were also directed to request Clearview to delete the information that it uploaded to the company's servers.

Clearview under scrutiny

US-based publication Buzzfeed approached the NBI last March to inquire if Finnish authorities had used the controversial app. The NBI denied that Clearview AI had been used in Finland.

About a week later, Buzzfeed told the NBI that, according to their information, the Finnish police had used the software about 120 times. The Police Board and the NBI immediately launched an investigation and found that Buzzfeed's information turned out to be correct.

The NBI informed the Data Protection Ombudsman about the app's use, saying the agency would stop using it.

Last week, online magazine Politico reported that Clearview AI had subpoenaed documents (siirryt toiseen palveluun) from organisations that had exposed the company's controversial activities.

According to Politico, the subpoenas could cause a chilling effect on other groups from reporting about Clearview AI or other firms working on "potentially problematic technologies."