An internal petition calling for Google to stay out of "the business of war" was gaining support Tuesday, with some workers reportedly quitting to protest a collaboration with the US military.

About 4,000 Google employees were said to have signed a petition that began circulating about three months ago urging the internet giant to refrain from using artificial intelligence to make US military drones better at recognizing what they are monitoring.

Tech news website Gizmodo reported this week that about a dozen Google employees are quitting in an ethical stand.

The California-based company did not immediately respond to inquiries about what was referred to as Project Maven, which reportedly uses machine learning and engineering talent to distinguish people and objects in drone videos for the Defense Department.

"We believe that Google should not be in the business of war," the petition reads, according to copies posted online.

"Therefore, we ask that Project Maven be cancelled, and that Google draft, publicize and enforce a clear policy stating that neither Google nor its contractors will ever build warfare technology."

– 'Step away' from killer drones –

The Electronic Frontier Foundation, an internet rights group, and the International Committee for Robot Arms Control (ICRAC) were among those who have weighed in with support.

While reports indicated that artificial intelligence findings would be reviewed by human analysts, the technology could pave the way for automated targeting systems on armed drones, ICRAC reasoned in an open letter of support to Google employees against the project.

"As military commanders come to see the object recognition algorithms as reliable, it will be tempting to attenuate or even remove human review and oversight for these systems," ICRAC said in the letter.

"We are then just a short step away from authorizing autonomous drones to kill automatically, without human supervision or meaningful human control."

Google has gone on the record saying that its work to improve machines' ability to recognize objects is not for offensive uses, but published documents show a "murkier" picture, the EFF's Cindy Cohn and Peter Eckersley said in an online post last month.

"If our reading of the public record is correct, systems that Google is supporting or building would flag people or objects seen by drones for human review, and in some cases this would lead to subsequent missile strikes on those people or objects," said Cohn and Eckersley.

"Those are hefty ethical stakes, even with humans in the loop further along the 'kill chain.'"

The EFF and others welcomed internal Google debate, stressing the need for moral and ethical frameworks regarding the use of artificial intelligence in weaponry.

"The use of AI in weapons systems is a crucially important topic and one that deserves an international public discussion and likely some international agreements to ensure global safety," Cohn and Eckersley said.

"Companies like Google, as well as their counterparts around the world, must consider the consequences and demand real accountability and standards of behavior from the military agencies that seek their expertise — and from themselves."

Twitter tweak steps up fight against trolls
San Francisco (AFP) May 15, 2018 –

Twitter said Tuesday it was stepping up its long-running battle against online trolls, trying to find offenders by looking at "behavioral signals."

The new approach looks at behavioral patterns of users in addition to the content of the tweets, allowing Twitter to find and mute online bullies and trolls.

Even if the offending tweets are not a violation of Twitter policy, they may be hidden from users if they are deemed to "distort" the conversation, Twitter said.

The announcement is the latest "safety" initiative by Twitter, which is seeking to filter out offensive speech while remaining an open platform.

Twitter already uses artificial intelligence and machine learning in this effort but the latest initiative aims to do more by focusing on the actions of certain users in addition to the content.

"Our ultimate goal is to encourage more free and open conversation," chief executive Jack Dorsey said.

"To do that we need to significantly reduce the ability to game and skew our systems. Looking at behavior, not content, is the best way to do that."

A Twitter blog post said the move aims at "troll-like behavior" which targets certain users and tweets with derisive responses.

"Some troll-like behavior is fun, good and humorous. What we're talking about today are troll-like behaviors that distort and detract from the public conversation on Twitter," said the blog from Twitter executives Del Harvey and David Gasca.

"Some of these accounts and tweets violate our policies, and, in those cases, we take action on them. Others don't but are behaving in ways that distort the conversation."

Harvey and Gasca said the challenge has been to address "disruptive behaviors that do not violate our policies but negatively impact the health of the conversation."

The new approach does not wait for people who use Twitter to report potential issues.

"There are many new signals we're taking in, most of which are not visible externally," the blog post said.

"Just a few examples include if an account has not confirmed their email address, if the same person signs up for multiple accounts simultaneously, accounts that repeatedly tweet and mention accounts that don't follow them, or behavior that might indicate a coordinated attack."

In some cases, if the content is not a violation of Twitter policies, it will not be deleted but only shown when a user clicks on "show more replies."

"The result is that people contributing to the healthy conversation will be more visible in conversations and search," Harvey and Gasca wrote.

Twitter said its tests of this approach shows a four percent drop in abuse reports from search and eight percent fewer abuse reports from conversations.