A Complete Tech News Blog!

Meta continues to implement the recommendations of the civil rights audit conducted last year

Meta

Meta says it is still working on a number of changes recommended by auditors more than a year after failing its first civil rights audit. Detailed progress was made on addressing many of the recommendations made by the auditors in an update released by the company.

A company statement indicates that 65 of the 117 recommendations have been implemented, and another 42 are in the process of being addressed.

As of now, the company says it is assessing six areas for “feasibility” of making changes and has “declined” to take any further action on two others.

In particular, some of the recommendations address some of the most contentious issues identified in the initial 2020 audit.

According to that report, released in July of 2020, Facebook needs to do more to stop “pushing users toward extremist echo chambers.”

It criticized the company’s handling of Donald Trump’s posts, as well as issues related to algorithmic bias.

According to its latest update, Meta still hasn’t made all of the algorithmic bias corrections outlined by auditors.

Some changes have been made, such as engaging outside experts and diversifying the company’s AI team, but others are still in the evaluation stage.

Meta

It recommends a mandatory, company-wide process for avoiding, identifying, and addressing possible sources of bias and discriminatory outcomes when developing or deploying AI and machine learning models, along with regular testing of algorithms and models already available.

Meta indicates that Meta is “under evaluation.” Similarly, the audit recommends “mandatory training on understanding and mitigating sources of bias in AI for all teams building algorithms and machine-learning models.” That recommendation is also “under evaluation,” Meta states.

It also mentions that some updates related to content moderation are under evaluation.

The recommendations include improving the “transparency and consistency” of decision-making related to moderation appeals, and studying more aspects of hate speech’s spread, and how it can use the data to more efficiently address targeted hate.

Meta was also recommended by the auditors to “disclose additional data” about which users are being targeted with voter suppression. A decision on that recommendation has not yet been made.

Meta declined only two recommendations that had directly to do with elections and census policies. Meta advised reviewing content reviewers to determine whether user-generated reports of voter interference violate our policies, and to add an appeals option for reported voter interference content.

However, the company decided against making those changes since it would slow down the review process, and because “the vast majority of content reported as voter interference does not violate the company’s policies.”

The company is also developing a framework to study its platforms and identify opportunities to improve fairness regarding race in the United States. To achieve this, Meta will conduct “off-platform surveys” and analyze its own data using surnames and zip codes.

Share on facebook
Share on twitter
Share on linkedin
Share on pinterest
Share on tumblr
Share on vk
Share on mix
Share on reddit
Share on email

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Articles

baseread

Content Creator

Baes Read brings all the most recent Consumer Technology achievements & demonstrates to you what’s going on, what makes a difference & how technology can enhance your life.

recent post