Microsoft removes a database containing more than ten million faces

(June 7)Earlier in the week Microsoft removed a database of more than ten million faces. The database was used to test and train facial recognition algorithms. The database was known as MS Celeb.

MS Celeb

The database consisted of more than 10 million images of about 100,000 people. The images were largely scraped from publicly available sources on line.The structured data that accompanied the photos made the dataset quite useful in training AI to recognize a persons face across several different photos.

Many in the photos were not aware that the data set existed

However a Financial Times Investigation found that many of the persons in the images were not aware of its existence and had not consented to their photos being used. A number of experts had speculated that the dataset could encounter legal issues. General Data Protection Regulation imposes significant requirement as to how a person's individual data is stored and transferred.

Microsoft does not announce removal of the dataset

Microsoft downplayed the significance of the site saying that it was intended for academic purposes and was run by an employee who was no longer with the company as he has been removed. Two similar data sets at DUke University and Stanford were also removed after the Financial Post investigation.Two similar data sets run by Duke University and Stanford were also taken down in the wake of the FT’s reporting.

The dataset was widely used

A recent article notes: " The images were gathered from search engines and published in 2016 to a dataset called MS Celeb and used to train facial recognition systems around the world, including by military researchers and Chinese firms such as SenseTime and Megvii, the Financial Times reported Thursday. The dataset — previously used in an AI project to recognize celebrities — had been linked to China’s efforts to crack down on ethnic minorities in the country."

The facial-recognition technology is used for everyday tasks such as unlocking phones and tagging friends on social media. However privacy concerns persist Law enforcement agencies also use the technology to help with investigations. However, the technology can often be flawed. As the appended video notes the FBI has a huge database of facial recognition images.

Previously published in the Digital  Journal

Facial recognition software generates 28 false positives linking US Congress members to mugshots

The American Civil Liberties Union (ACLU) tested Amazon's facial recognition system by scanning all 535 members of the US Congress against 25,000 mugshots public mugshots. The test generated 28 false positives.

	1 of 3
None of the members were in mugshots. The Amazon system called Rekognition is already in use by some police departments.

The ACLU wants a moratorium on the use of Rekognition by police

ACLU said: “An identification — whether accurate or not — could cost people their freedom or even their lives. Congress must take these threats seriously, hit the brakes, and enact a moratorium on law enforcement use of face recognition.”

Amazon blames results on faulty calibration

An Amazon spokesperson said to Verge that the result was due to poor calibration. The ACLU tests were done using Amazon's default setting of 80 percent. However, the spokesperson said that for law enforcement applications at least a 95 percent application should be used as a false ID could have significant consequences.

The Amazon representative said: “80% confidence is an acceptable threshold for photos of hot dogs, chairs, animals, or other social media use cases. It wouldn’t be appropriate for identifying individuals with a reasonable level of certainty. it wouldn’t be appropriate for identifying individuals with a reasonable level of certainty.”

Yet, there is nothing in the setup process that enforces the recommendation meaning that there is nothing to prevent police from using the default setting.

Amazon reported a number of police depts. used Rekognition

The Orlando police in Florida used the system as a pilot project in real-time recognition. The system is sold as part of Amazon's Web Services. It can often cost less than $12 a month for an entire department. Although the Orlando project has ceased, the department continues to use the system.

ACLU is examining Recognition use by an Oregon Sheriff's Department

The latest ACLU experiment is designed in particular to examine Washington County's Sheriff's Department in Oregon's use of Recognition. The department has compared images against a database of as many as 300,000 mug shots.

Jacob Snow who organized the test for the ACLU of Northern California said: “It’s not hypothetical. This is a situation where Rekognition is already being used.”

Test shows signs of racial bias

11 of 28 or about 39 percent of false matches were of people of color and included civil rights leader Rep, John Lewis a Democrat of Georgia and five other members of the Congressional Black Caucus. Only 20 percent of the present Congress are people of color. People of color were obviously identified at a higher rate.

Problems with the use of Rekognition

In practice many facial recognition positives would be confirmed by a human before they led to concrete action such as an arrest. However, Snow said: “Imagine a police officer getting a false match for somebody with a concealed weapon arrest. There’s a real danger if that information is surfaced to the officer during a stop. It’s not hard to imagine it turning violent.”

Snow told Reuters:“Face surveillance is flawed, and it’s biased, and it’s dangerous.”

Nevertheless the outlook of Amazon was positive and touted a range of uses for Rekognition according to a spokeswoman for Amazon Web Services:“We remain excited about how image and video analysis can be a driver for good in the world.” She said the system could be used to help find lost children and preventing crimes. She noted that they system was usually used to narrow the field rather than make final decisions.

Amazon workers protest sales of facial recognition software to the police

Workers at Amazon are demanding that the company cease selling facial recognition software to the U.S. government as it would be contributing tools that violate human rights.

	1 of 2

A letter to CEO Jeff Bezos cites recent mistreatment of refugees and immigrants by ICE as well as the targeting of black activists by law enforcement. The letter noted as a parallel IBM's sale of computer services to the Nazis: “As ethically concerned Amazonians, we demand a choice in what we build, and a say in how it is used. IBM did not take responsibility then, and by the time their role was understood, it was too late. We will not let that happen again.”

Amazon Web Services Rekognition and AWS cloud services

The workers said that Amazon Web Services Rekognition, the facial recognition software, should not be sold to law enforcement. The AWS cloud services should not be sold to Palantir. Palantir is a data analytics firm that provides mission critical software to Immigration and Customs Enforcement (ICE).

The sale of Rekognition software to the police was discovered by a American Civil Liberties Union (ACLU) investigation in May. The ACLU warned that the deployment of such technology could be the beginning of automated mass surveillance in the US.

Palantir has been working with ICE since 2014 when Obama was president. It helps the agency manage the mass of personal data needed to target and deport individuals. The letter makes particular reference to the separation of families at the US border as a prime motivation for their protest. The entire letter can be found here.

The letter says about Palantir: "We also know that Palantir runs on AWS. And we know that ICE relies on Palantir to power its detention and deportation programs. Along with much of the world we watched in horror recently as U.S. authorities tore children away from their parents. Since April 19, 2018 the Department of Homeland Security has sent nearly 2,000 children to mass detention centers. This treatment goes against U.N. Refugee Agency guidelines that say children have the right to remain united with their parents, and that asylum-seekers have a legal right to claim asylum. In the face of this immoral U.S. policy, and the U.S.’s increasingly inhumane treatment of refugees and immigrants beyond this specific policy, we are deeply concerned that Amazon is implicated, providing infrastructure and services that enable ICE and DHS."

We will have to wait and see whether the protests at Amazon will change policy but the trend is for tech workers to take an active stance when their companies engage in work which is questioned by many including employees.

The Amazon protest follows earlier ones at Google and Microsoft

In March, it was revealed that Silicon Valley workers were helping build AI tools to analyze drone surveillance footage. In response workers protested and more than a dozen even resigned. The protest was successful and Google pulled out of the contract and even pledged not to develop AI weapons.

Google had offered its resources to the US Department of Defense for Project Maven. This is a research initiative to develop computer vision algorithms that can analyze drone footage. More than 3,100 Google employees signed a letter asking Sundar Pichai the Google CEO to reevaluate the company's involvement. The letter said that Google should not be involved in the business of war.

Microsoft has attempted to downplay its work with ICE but this has not stopped internal dissent. More than 300 employees have signed a letter denouncing the campaign. Microsoft CEO Satya Nadella claims the company was only providing benign software for tasks such as messaging and email. It remains to be seen whether the protests will result in any change of policy.

Previously published in Digital Journal