Header Ads

Safety as a service

Hi, it's Mark Bergen. Google's former head of Trust and Safety has a few unorthodox ideas: Create a government agency to monitor social media companies as if they were polluters. Tax internet platforms that host user-generated content. Stop letting technology giants make content moderation decisions—and instead hand over moderation duties to independent auditors.

Those are some of the new proposals published on Wednesday by Trust Lab, a startup pitching itself as one such outside moderation shop for tech platforms. Trust Lab was co-founded by Tom Siegel, who spent 14 years at Alphabet Inc.'s Google, including a stint running the division that handles enforcing rules on content policies for its various properties.

Siegel's Trust Lab is part of a new crop of companies and organizations trying to come up with automated solutions to online abuse. Others include Hive, Sentropy and Spectrum Labs. To borrow from tech parlance, think of their offerings as trust-and-safety-as-a-service. "If users have an opportunity to contribute content in any form, abuse will take place," said Rahul Roy-Chowdhury, another former Google executive and an investor in Trust Lab. "The opportunity here is massive."

Siegel left Google in early 2019. Since then, he's watched his alma mater, Facebook Inc. and other platforms heave more money into moderation, and take a shellacking for how they've managed misinformation, hate speech and nefarious actors. "Everyone agrees," said Siegel, "it's a real mess."

The current method for social platforms disclosing their moderation work—in which each company voluntarily reports different metrics for content removal, with scant details—is "meaningless," according to Siegel, and in need of reform. He thinks outside firms should be licensed to run these disclosure reports and to perform platform cleanups. He also wants a federal oversight body to help set rules, envisioning a new government entity as a sort of "Internet Protection Agency," modeled after the Environmental Protection Agency and the other U.S. organizations that police consumer safety and finance.

"This common sense approach is already implemented in almost every other sector of society where harm occurs," Siegel wrote in Wednesday's post.

When critics slam big tech companies for lax moderation, the proposed solutions most often focus on changes to U.S. laws that insulate platforms from liability for content on their sites. But Siegel, a technologist, is most concerned with reforming operations. Trust Lab provides software to analyze and track unwanted online material. 

Often, tech companies contract out much of their human moderation operations to outfits like Accenture PLC and Vaco—though those companies have a checkered history of worker treatment, and typically just follow company direction or augment work done by artificial intelligence systems. Siegel and his investors believe there's an opportunity for startups. Neil Daswani, a cybersecurity expert and a Trust Lab backer, said such companies represent a growing but overlooked trend. They're pitching fixes to both huge platforms and smaller ones that must arbitrate online crowds with fewer resources. 

Trust Lab declined to share its clients, but Siegel said it's working with "leading social media companies, marketplaces and messaging apps."

Of course, Siegel's most sweeping proposals could only work if both Washington and Silicon Valley sign on. Tech giants have openly invited more regulations around policing speech. But these same companies are reluctant to open up their own rules books and algorithms, the kind of practice that would be required for a standardized, audited moderation system. While platforms have coordinated to combat child pornography and terrorism, they haven't on misinformation or hate speech, despite calls to do so.

The closest step in this direction might be the Facebook Oversight Board, which has received hundreds of thousands of complaints from Facebook users since last fall. "It's a really good attempt," Siegel said. "But I don't think it's going to be enough." —Mark Bergen

If you read one thing

Bitcoin dipped below $30,000 for the first time since January on Tuesday, before ticking back upward and recouping the day's losses. The wild volatility follows China's cryptocurrency crackdown.

And here's what you need to know in global technology news

Amazon and Walmart are making peace with kiranas, the ubiquitous local shops in India that account for 90% of the country's purchases.

A new European Union ad tech probe strikes at the core of Google's business model.

Amazon's takeover of movie studio Metro-Goldwyn-Mayer will be reviewed by the U.S. Federal Trade Commission, which is now being led by prominent Amazon critic Lina Khan.

Marc Andreessen holds forth on techno-optimism, software eating the world and more.  

It's almost time to Power On. A new weekly newsletter by Bloomberg's Mark Gurman will deliver Apple scoops, consumer tech news, product reviews and the occasional basketball take. Sign up to get Power On in your inbox on Sundays.

 

Like Fully Charged? | Get unlimited access to Bloomberg.com, where you'll find trusted, data-based journalism in 120 countries around the world and expert analysis from exclusive daily newsletters.

No comments