❓Microsoft, Amazon and IBM announced they’re pausing selling their facial recognition technology to police following evidence showing inherent racial bias. Privacy International asks whether IBM is genuine,  Amazon’s hypocrisy has been flagged, and Azeem Azhar’s Exponential View lists the questions we should ask now. 

2 good podcast episodes:

1. Is this the end of facial recognition? (What next, TBD), with guest Deb Raji, technology fellow who describes the lack of diversity in face datasets as well as her colleagues’ reluctance to think about an extra dimension of representation when it is already “so hard to collect data”. Raji previously teamed up with Joy Buolamwini, founder of Algorithmic Justice League whose Coded Bias documentary is being screened online on 20 June). 

2. Amazon finally sees the problem with facial recognition (Intelligencer, NY Magazine), Tech journalist Kara Swisher and Professor Scott Galloway discuss tech companies’ responsibility for the issues. 
    

🇵🇭Unease in the Philippines
😐 Creepy: many Facebook accounts belonging to critics of the Philippines’ President Rodrigo Duterte have been cloned. 😬 Sinister: some duplicates have been messaging the real owners with death threats.
This piece asks who’s behind the surge in the fake accounts? Remember in 2016 when Cambridge Analytica bragged they could get a "fundamentally flawed" candidate elected easily? And then Duterte was elected. Is someone beginning to flex, flaunt and show off ahead of the 2022 election?
 

✅ Journalist James Ball explains why simply breaking up big tech monopolies won't solve anything. We need an “entirely new social contract [because] this isn’t just about more privacy protections: it’s about rights to our data, audits of algorithmic decision-making, registers of use, and new tax systems for a world shrunk by the web.”
 

💳 And this is fun. Peter from Built for Mars pits traditional and challenger banks against each other in this series of posts. “Pick a winner between Monzo, Revolut and Starling?” he says. He will in later chapters.

👏🏿 Watch and learn

Interview with Aurdey Tang, Taiwanese Digital Minister, on how Taiwan – a country with 24 million people and regular travel to China – became a coronavirus success story (450 cases and 7 deaths at the time of recording). This is how it’s done.

🇯🇵 Accelerating digital transformation 

Coronavirus has forced the Japanese health ministry to improve an outdated paper-and-fax system that had “been under debate for at least 10 years” before it was called out for inefficiency “in a war-zone-like situation where the healthcare system is about to collapse." 

Open, standardised, accessible (ethnicity) data

🇨🇦 Superb collaboration between California’s Department of Public Health and the Department of Technology. Clearly annotated breakdown of coronavirus data in the state of California. The ‘who is getting infected’ section reveals significant disparities between ethnicities and includes ‘structural racism’ as a possible explanation of why, as well as poverty and an increased likelihood of having underlying conditions. 

🇬🇧 Meanwhile in the UK, there’s no one source of truth. 

• The ONS published an online report which includes about 5 weeks’ worth of data.

• Public Health England have published a PDF review showing a “disproportionate impact on Black, Asian and minority ethnic groups”. There’s proposed action but it doesn’t equal *actual* action... ¯\_(ツ)_/¯

• The NHS has been publishing daily deaths since 2 April (ethnicity breakdown included in weekly updates), while the BBC’s once prominent daily deaths data has vanished from the homepage. 

The Ethnicity facts and figures site (sole purpose to highlight disparities between the experience and treatment of different ethnic groups) – is oddly silent. There’s not a single mention of the virus on there. Why? 😶