It may be chilly now in Google’s hometown of Mountain View, California. But that’s nothing compared to the frigid temperatures expected on Tuesday in Washington, D.C., when Google CEO Sundar Pichai heads to Capitol Hill. Having skipped a Senate Intelligence Committee panel in September on state-sponsored election meddling on technology platforms, Google will be the sole guest at a House Judiciary Committee hearing (at 10 a.m. ET) titled, “Transparency & Accountability: Examining Google and its Data Collection, Use and Filtering Practices.” Pichai’s grilling will include charges of search result political bias, mishandling of user security on the moribund Google+ social network (which just revealed another leak), and collaboration with China’s repressive regime on the Dragonfly search engine project.
Republican representatives will likely bring up charges of bias against conservative voices in Google search results. They will have a hard time proving their point, since good search results are intended to surface valuable information–not the preferred conspiracy theories of far-right sites like Gateway Pundit—and since it doesn’t make sense for Google, which makes money by surfacing ads on results pages, to deliberately alienate people of a particular political view.
But don’t expect a nuanced, informative discussion of algorithm bias, which is likely beyond the technical understanding of most representatives—so look forward to tough questions from Republicans and platitude-filled replies from Pichai describing the company’s commitment to fairness and objectivity.
Likely more interesting will be the topic on which both Republicans and Democrats have some common ground: concern over Google’s well-reported Dragonfly project to build a search engine product for China that not only censors results but tracks what citizens search for.
Mistrust of China spans the political spectrum, due to geopolitical and human rights concerns. Republicans may be especially incensed that Google is pulling out of supplying AI technology to the Pentagon’s Project Maven, while continuing to pursue a deal with the country’s military and economic rival. Human rights concerns have been especially inflamed by the widely reported detention of 800,000 to 2 million Uighurs and other Muslim minorities in concentration camps.
Google’s only defense on Dragonfly so far is that the project is in its initial stages–something contradicted in extensive reporting by the Intercept. And even if Google’s assertions were true, it’s not much of a defense.
After withdrawing from search in China in 2010 over ethical concerns and a hacking attack, and after publishing a statement of AI principles in response to employee protests over Maven, Google appears to have ignored the high-minded statements it made in both those cases. So says scientist Jack Poulson, who resigned from Google in opposition to Dragonfly last August. He cites in particular what he claims is Google’s hypocrisy over its pledge to not “design or deploy . . . technologies whose purpose contravenes widely accepted principles of international law and human rights.”
“I think what we should look at is their actions, not their words,” Poulson said on a teleconference this morning, convened by a new tech and human-rights coalition opposing Dragonfly. “They very much would like to promote these [AI] principles as something they stand for. But when it comes time for actually making a tough decision and needing to stand on these principles in a difficult way, there’s a refusal.”
It’s a long shot. But if members of Congress can get Pichai to go beyond platitudes and talk about how lofty principles do or don’t apply to on-the-ground decisions, take note.