Connect with us

Big Tech

House Republicans Grill Google CEO Sundar Pichai Over Alleged Political Bias

Published

on

WASHINGTON, December 11, 2018 – The CEO of search engine Google came to Washington on Tuesday and politely rebutted all charges that the world’s largest search engine is biased against conservative viewpoints.

In the calm and controlled voice of an engineer, CEO Sundar Pichai said, “Our products are built without any bias,” responding to a question of the House Judiciary Committee Chairman.

“We don’t build partisan features,” he repeated later to another Republican member of the committee.

Indeed, Pichai had to contend with an almost-uninterrupted narrative – fed largely, but not completely, by members of the GOP – that Google’s search engine results were in some way systematically biased.

Moreover, as the head of one of the country’s leading information technology companies, Pichai was robustly challenged on issues ranging from the extent of Google’s surveillance-like data-collection to the existence of prototype search engine that returned restricted results – and is apparently designed to cater to the communist China market.

Pichai didn’t face as much overt hostility as was experienced by Facebook CEO Mark Zuckerberg when he sat down for his grilling before the same committee after the Cambridge Analytica data privacy scandal earlier this year. But Pichai wasn’t welcomed very warmly, either.

On privacy, Pichai said that he, like Facebook CEO Zuckerberg, supported Congress considering data privacy legislation.

Does Google exhibit bias in its search engine results?

Committee Chairman Bob Goodlatte, R-Va., set a confrontational tone, without being conspiratorial:

“While it is true that Google is not a government entity and so it does not have to comply with the First Amendment, the American people deserve to know what types of information they are not getting when they perform searches on the internet. The market works best when information about products and services is readily available, and so today – on behalf of this Committee and the American consumer – I hope to get answers from Mr. Pichai regarding who at Google makes the judgment calls on whether to filter or block objectionable content and what metrics Google uses to make those decisions.”

Pichai insisted that Google’s algorithms are designed to accurately reflect what people are talking about online at any given time. Google is not “the internet” so much as representing what is on the internet at any given time, he seemed to be saying.

“Any time you type in a keyword, we crawl copies of billions of web pages, and we take the keyword and match it against pages for relevance, freshness, popularity, how others are using the it, and we try to rank and find the best” pages for that particular keyword, he said.

Rep. Zoe Lofgren, D-Calif., used this concept to explain why, when an individual conducts a search for the word “idiot” in Google, the image of Donald Trump comes up repeatedly.

Google doesn’t return these results because Google is making this commentary of the president, Lofgren said. Rather, Google is reflected what others internet users are saying.

Rep. Ted Lieu, D-Calif., made the same point when noting that search results of most of his fellow colleagues were not overtly imbalanced – with the exception of Rep. Steve King, R-Iowa (not a member of the Judiciary Committee) – but has recently made controversial statements about figures linked with white supremacy movements.

“If you are getting bad search results on Google, don’t blame Google, blame yourself,” Lieu said to his Republican colleagues.

“This is the fourth hearing in a series of ridiculous hearings [because] the First Amendment protects private individuals and corporations’ rights to freedom of speech.”

Republican representatives pile on against Google

Still, Republican after Republican had a story to tell about a gripe they had regarding Google results. Many lobbed in questions about Google’s privacy and market power.

Rep. Lamar Smith, R-Texas, refused to believe that bias was not present in the curation of political content when more than 90 percent of the searches for Donald Trump produce negative stories on the president. He also referred to pro-Trump content being labelled, or “flagged,” as hate speech.

“This doesn’t happen by accident, but is baked into the algorithms,” said Smith.

Pichai disagreed, and referred to the company’s political neutrality in algorithm results as “sacrosanct.”

But Smith wasn’t buying it, and referred to the evidence of political bias on the Google platform as “irrefutable.”

Rep. Steve Chabot, R-Ohio, contended that virtually every reference to a health care bill that he had introduced was “an attack on our bill.” He had to go to the third or fourth page of the search engine’s results to find one that was “remotely positive.”

Replied Pichai: “We use a methodology about what is being said about a topic at any given time.

“It is in our interest to make sure we reflect what is happening out there in the most effective method possible. Our algorithms have no sense of politics.”

Rep. Darrell Issa, R-Calif., criticized an apparent divergence in the rates that are charged for the keywords being used by Republican candidates versus Democratic candidates.

Pichai said that prices for advertising were determined on the basis of automatic auctions, and that that was “why I am confident that we don’t approach our work with political bias.”

Nonethless, Pichai committed to following up with Issa to looking at the reasons for the divergent pricing of keywords for Republican versus Democratic candidates.

Rep. Ted Poe, R-Texas, pressed for details on the information collection capacities of an Android phone. Then, acknowledging that Google had a First Amendment right to present the search results it wanted, added: “I hope we don’t get to the point where government comes in and regulates what is biased, because [Google] is an independent and free company.”

Privacy and China also play a role in the hearing

Pichai was also criticized repeatedly – by representatives of both parties – for its sweeping data-collection practices, and for a reported prototype of a search engine for the Chinese market.

Pichai wiggled on the question of a Chinese search engine: “We have no plans to launch in China,” he said, adding, “Right now, we have no plans to launch search in China.”

Rep. Keith Rothfus, R-Penn., finally got the most information out of him, when he acknowledged the existence of a prototype products for “what search would look like” in a country that require mandatory content filtering.

At one point, he said, more than 100 Google engineers were working on the project.

(Photo of Google CEO Sundar Pichai by Drew Clark.)

 

Big Tech

Tech Policy Conference Panelists Tackle Challenges of Federal Privacy, Antitrust Laws

Academics were concerned about an anti-preference bill, while one state AG said he’s ‘pragmatic’ about a federal privacy law.

Published

on

Screenshot of Colorado Attorney General Phil Weiser at the TPI Aspen Forum on Monday

ASPEN, Colorado, August 15, 2022 – Academics expressed concern Monday about antitrust legislation before Congress that would prevent companies from preferencing their own products on their platforms, arguing the legislation targets only certain companies and hasn’t shown it would benefit consumers.

The American Innovation and Choice Online Act, S.2992, which is currently before the Senate and aims to ban discrimination against third-party products on the host platform, defines targeted companies by their value – which effectively narrows the number of affected companies and makes it a problematic piece of legislation, according to some academics.

“I think it’s very difficult to single out specific companies…for specific rules,” Judy Chevalier, a professor of finance and economics at Yale University, said at the TPI Aspen Forum on Monday.

“It’s hard to imagine what is the principle whereby private label band aids are a bad idea at Amazon but they’re a good idea at Walmart,” she added. “The self-preferencing rule can be applied to Amazon in a way that I think can be interpreted to limit their ability to introduce and promote their private label products.

“It’s not very convincing that this behavior has thus far harmed consumers,” she continued. “So I think singling out particular companies in this broad brush way strikes me as problematic.”

Dennis Carlton, a professor of economics at the University of Chicago business school, said the legislation makes him “nervous” because of the impact on innovation of targeting certain industries over others.

“High tech industries are rapidly changing, and whenever we have regulation or try and have regulation of rapidly changing industries, it is just too hard for the regulators to keep track of what’s going on and they wind up causing delays in innovation,” Carlton said.

“Innovation is one of the strongest ways we improve our products and our standard of living. It makes me very nervous when you target specifically an industry or…make exceptions to other industries without…economic criteria or any attempt to show that this would produce a benefit not a harm. So it makes me nervous these proposals.”

Similar sentiments were expressed on a Broadband Breakfast panel in March, in which an association representing large technology companies blasted the legislation introduced by Senator Amy Klobuchar, D-Minn., as unfairly targeting certain online platforms and excluding large retailers.

“The bill very carefully picks winners and losers,” said Arthur Sidney, vice president of public policy at Computer and Communications Industry Association, which includes members like Amazon, Google, and Facebook.

State AGs weigh in on privacy legislation

On a separate panel at the Forum on Monday, the state attorneys general of Colorado and Nebraska discussed the state of privacy legislation – both in their own state and at the federal level.

Introduced in June, the American Data Privacy and Protection Act (H.R. 8152) cleared the House Energy and Commerce Committee last month for House floor votes. The proposed bill would provide Americans protections against discriminatory use of their data, require covered entities to minimize the data they collect, and prevent customers from needing to pay for privacy.

Despite his state having passed comprehensive privacy laws that are considered leading and a model by some, Colorado AG Phil Weiser said he’s “pragmatic” about a federal law.

“If a federal law is as good and strong as what we worked on in Colorado, I am comfortable with that law preempting Colorado, provided state AGs have the authority to enforce federal law,” he said. “It’s important to me to have that model because, you could imagine a world where the feds are not engaged in active enforcement, then the states can pick up that slack.”

Before the introduction of the legislation, some experts were concerned that having a number of different state privacy laws would harm smaller companies operating across multiple states. One lawyer noted that the longer companies have to wait for a uniform federal law, the greater the burden of compliance on them.

In fact, two Democratic California reps – Anna Eshoo and Nanette Barragan – were concerned that such a federal law would override their own state’s law. Eshoo proposed a provision, which was not included during a markup of the bill, that would have allowed states to add privacy provisions on top of the federal baseline.

“If you do have multiple standards,” Weiser said, “we have to solve for the problem, which is a problem right now of what I call interoperability or harmonization: How do we make sure that different state laws enable compliance across them as opposed to putting businesses in, to me, the unacceptable position of saying, ‘I can either comply with Colorado’s law or California’s law, but not both.’?”

Having had a privacy proposal in its legislature that did not pass, Doug Peterson, AG for Nebraska, said the state is taking a wait-and-see approach, including observing how states, including Colorado, fare with their own laws.

Continue Reading

Social Media

Americans Should Look to Filtration Software to Block Harmful Content from View, Event Hears

One professor said it is the only way to solve the harmful content problem without encroaching on free speech rights.

Published

on

Photo of Adam Neufeld of Anti-Defamation League, Steve Delbianco of NetChoice, Barak Richman of Duke University, Shannon McGregor of University of North Carolina (left to right)

WASHINGTON, July 21, 2022 – Researchers at an Internet Governance Forum event Thursday recommended the use of third-party software that filters out harmful content on the internet, in an effort to combat what they say are social media algorithms that feed them content they don’t want to see.

Users of social media sites often don’t know what algorithms are filtering the information they consume, said Steve DelBianco, CEO of NetChoice, a trade association that represents the technology industry. Most algorithms function to maximize user engagement by manipulating their emotions, which is particularly worrisome, he said.

But third-party software, such as Sightengine and Amazon’s Rekognition – which moderate what users see by bypassing images and videos that the user selects as objectionable – could act in place of other solutions to tackle disinformation and hate speech, said Barak Richman, professor of law and business at Duke University.

Richman argued that this “middleware technology” is the only way to solve this universal problem without encroaching on free speech rights. He suggested Americans in these technologies – that would be supported by popular platforms including Facebook, Google, and TikTok – to create the buffer between harmful algorithms and the user.

Such technologies already exist in limited applications that offer less personalization and accuracy in filtering, said Richman. But the market demand needs to increase to support innovation and expansion in this area.

Americans across party lines believe that there is a problem with disinformation and hate speech, but disagree on the solution, added fellow panelist Shannon McGregor, senior researcher at the Center for Information, Technology, and Public Life at the University of North Carolina.

The conversation comes as debate continues regarding Section 230, a provision in the Communications Decency Act that protects technology platforms from being liable for content their users post. Some say Section 230 only protects “neutral platforms,” while others claim it allows powerful companies to ignore user harm. Experts in the space disagree on the responsibility of tech companies to moderate content on their platforms.

Continue Reading

Big Tech

Surveillance Capitalism a Symptom of Web-Dependent Companies, Not Ownership

Former Google executive Richard Whitt critiqued Ben Tarnoff’s argument in ‘Internet for the People’ during Gigabit Libraries discussion.

Published

on

Photo of Ben Tarnoff, co-founder of magazine Logic and the author of “Internet for the People”

July 15, 2022 – A former Google executive  pushed back against a claim that the privatization of broadband infrastructure has created the world’s current data and privacy concerns, instead suggesting that it’s the companies that rely on the web that have helped fuel the problem.

Richard Whitt, president of technology non-profit GLIA Foundation and former employee of Google, argued that while the World Wide Web is rife with problems, the internet infrastructure underlying the web remains fundamentally sound.

Whitt was responding to claims made by Ben Tarnoff, a journalist and founder of Logic Magazine, at the Libraries in Response event on July 8. Tarnoff argued – as he does in his recent book, “Internet for the People” – that the privatization of broadband infrastructure in the 1990s has allowed the use and commodification of personal data for profit to flourish (known as surveillance capitalism).

The discussion took place during the Gigabit Libraries Network’s series “Libraries in Response.” The session was titled “If the Internet is Broken, How Can Libraries Help Fix it?”

Privatization, Tarnoff claims, has raised such issues as polarization of ideologies and the “annihilation of our privacy.” As a result, he said, the American people are losing trust in tech companies that “rule the internet.”

Whitt responded that the internet is working well based on the protocols, standardized rules for routing and addressing packets of data to travel across networks, derived at the onset of the internet.

The World Wide Web, a system built on the internet to allow communication using easy-to-understand graphical user interfaces, allowed for browsers and other applications to emerge, which have since perpetuated surveillance capitalism into the governing approach of the web that it is today, said Whitt, suggesting it’s not ownership of the hard infrastructure that’s the problem.

The advertising market that encourages surveillance extraction, analysis and manipulation is, and will continue to be, profitable, Whitt continued.

The discussion follows a Pew Research Center study that found that only half of Americans believe tech companies have a positive effect in 2019 compared to a seventy-one percent in 2015.

Continue Reading

Recent

Signup for Broadband Breakfast

Get twice-weekly Breakfast Media news alerts.
* = required field

Trending