U.S. Sens. Jeff Merkley (D-Ore.) and Bernie Sanders (I-Vt.) introduced legislation that would prohibit private companies from collecting biometric data—including faceprints and fingerprints—without consumers' and employees' consent. The National Biometric Information Privacy Act of 2020 would allow consumers and state attorney general to file lawsuits if their biometric data is obtained without permission. It would also bar companies from selling that data.

More:

• The proposed federal law mirrors a biometric privacy law in Illinois, which resulted in a $650m proposed settlement by Facebook for its photo-tagging feature.
• Merkley said the goal is to protect against a "'big brother' surveillance state" that threatens privacy and personal control of data.
• Sanders called facial recognition "Orwellian" and said the U.S. can't allow it to continue to violate Americans' civil liberties and privacies,
• The bill would bar companies like Clearview AI from providing facial recognition services outside of government contracts.
• Under the law, written consent couldn't be combined with an employment contract or "any other consent or permission-seeking instrument or function."

Microsoft could use video data from its potential acquisition of TikTok to train AI systems and learn more about its consumers. The Washington Post reports that Microsoft's lack of a video service makes it less competitive in the AI space compared to rivals like Facebook and Google.

More:

• If Microsoft were to acquire TikTok, it could harvest data in the form of billions of videos uploaded to TikTok every year, which could feed its AI ambitions.
• The data could train AI systems to understand better consumer trends, popular brands, and what makes a video go viral. It could also train computer systems to read lips.
• Amazon owns Twitch, Google owns YouTube, and Facebook has access to videos posted on its platform. Among large AI companies, Microsoft lacks this source of consumer video data.
• Microsoft is currently negotiating with TikTok parent company ByteDance to purchase the app’s operations in the U.S., Canada, Australia, and New Zealand. The company plans to reach a negotiation deal by Sept. 15th.
• President Trump will allow a TikTok sale to a U.S. company only if the U.S. government gets a cut of the deal. TikTok founder Zhang Yiming said that Trump's attempt to force a sale feels like it was made to ban the app.

Top podcasts in AI/ML

Next Wednesday, MIT Tech Review launches its highly-anticipated "In Machines We Trust" podcast. In honor of the series, we've decided to list our top picks for podcasts in the AI/ML space:

1) In Machines We Trust

Host: MIT Tech Review Senior Editor Jennifer Strong.

Topics: Automation, facial recognition, practical AI applications.

Notable: One upcoming podcast will interview a man who was wrongly arrested because of a faulty facial recognition match...

AI company Truera emerged from stealth to launch a platform that can identify and mitigate bias in AI models. The platform, which aims to fix machine learning's "black box" problems, is already being used by several Fortune 100 customers in banking and insurance.

More:

• Truera, previously known as AILens, was co-founded by Anupam Datta, Shayak Sen, and Will Uppington.
• The platform is based on explainability AI research done by Datta and Sen, both of Carnegie Mellon University. It grew out of research about gender bias in online advertising.
• The company's Model Intelligence platform aims to solve black-box problems like fairness, bias, and transparency in high-quality models.
• Truera also announced a $5.1m funding round led by Greylock with additional investors including Wing VC, Conversion Capital, and Aaref Hilaly. The cash will help it expand its service

A developer trained an AI model to convert people's photos into cartoons in the style of South Korean cartoonist Lee Mal-nyeon. A GitHub user by the name of bryandlee created a dataset of cartoon faces and trained a face generating model in just ten hours using 500 annotated images.

More:

• The developer incorporated FreezeD, the baseline for transfer learning of GANs, and U-GAT-IT, an image-to-image translation method.
• The code for FreezeG is on GitHub.
• Details on the training can be found on the malnyun_faces project page.
• A Reddit user said it was the first time they had seen when a network "actually understands what it is seeing and redraws it in a different style."

Toshiba plans to build an R&D AI center in Kawasaki, Japan. The $321m hub would accommodate around 3,000 employees who specialize in areas like AI, quantum cryptography, and security.

More:

• The center, which will have two buildings, is slated to open in April 2023. It will be rebuilt from an existing R&D center near Tokyo.
• Toshiba recently debuted its Motion Estimation AI, which can predict car and drone movement at high accuracy. The electronics and machinery maker also developed an AI system that can provide real-time video subtitles in online classes.
• Last year, Toshiba partnered with the University of Tokyo to train students in AI and big data analysis. Toshiba wants to triple its number of AI engineers to 2,000 by March 2023.

CSAIL researchers developed an AI system that can predict whether an AI or human is a better decision-maker. The system relies on two separate machine-learning models to make decisions, like diagnosing a patient, in tandem with humans. One model makes the decision or diagnosis, and the other predicts whether the AI should defer to a human.

More:

• The paper comes from Hussein Mozannar and David Sontag, who are researchers at CSAIL, MIT’s Computer Science and AI Laboratory
• Their decision-making system could apply not just to clinical applications like assessing chest X-rays, but the moderation of social media posts.
• The second model, known as the "rejector," can improve its predictions over time.
• The research is important as more medical environments use AI and automation to support human clinicians.

QUICK HITS

• Voiceitt, a commercial speech recognition platform, raised $10M in funding from Viking Maccabee Ventures, M12, and others.
• McAfee researchers tricked a modern facial recognition system into seeing a person who wasn’t there.
• Investors, including Alibaba, are adding an additional $300m into pre-IPO funding round for Chinese startup Xpeng Motors, which has autonomous driving technology.
• Men’s skincare is the latest industry to be disrupted. Read why times have changed.*

*This is a sponsored post.