When Twitter introduced voice tweets earlier this summer as a way to send more personalized messages, it caught a lot of flack for not including accessibility features. Now, the company said it will add transcriptions for voice tweets as part of an initiative to promote accessibility within its products. It’s also expanding the voice tweet feature, which is still only available on iOS, effective immediately.

We’re rolling out voice Tweets to more of you on iOS so we can keep learning about how people use audio.

Since introducing the feature in June, we’ve taken your feedback seriously and are working to have transcription available to make voice Tweets more accessible. (1/2)

— Twitter Support (@TwitterSupport) September 29, 2020

After voice tweets arrived in a testing phase, critics pointed out that it should have had captions from the get-go as required by Federal law. Twitter then surprisingly admitted that it didn’t have a dedicated accessibility team, but instead relied on employees to donate additional time to work on those features. Since then, the company has launched two separate teams dedicated to accessibility, including one for its products.

As an additional measure, Twitter started distributing phishing-resistant security keys to its employees and requiring its teams around the world to use them. Google implemented the measure in 2017 to great success: A year after making it mandatory for employees to use physical security keys for two-factor authentication, the tech giant announced that it has “no reported or confirmed account takeovers” anymore.

Twitter required all new employees to go through security, privacy and data protection trainings, as well. Those who have access to non-public data had to attend additional mandatory training sessions on how they can avoid becoming phishing targets for attackers. The company also said that it’s been constantly improving its internal detection and monitoring tools that alert the company of possible unauthorized access attempts.

As for its election-specific efforts, Twitter said it recently implemented heightened security measures for election-related Twitter accounts in the US. A few days ago, it started sending them in-app notifications on new security requirements going forward, such as enabling password reset protection for accounts by default. It also conducted additional penetration testing and scenario planning over the past months. From March 1st to August 1st, for instance, its cross-functional elections team performed exercises on how to deal with hacks, leaks of stole materials, foreign interference and coordinated online voter suppression campaigns, among other scenarios.

As a closer to its post, Twitter promised to roll out improvements to its privacy settings in the near future:

“We are continuing to invest more in the teams, technology, and resources to support this critical work. We also know that we can do more to make it easier for you to find and use the settings and controls we offer, so we’re working on rolling out improvements to the design and navigation of our privacy settings. You’ll see these improvements in Twitter soon.”

Twitter’s experiment that prompted people to read an article before tweeting it out has succeeded in getting more people to read before they tweet, the company said. 

The company has been testing the feature since June, when it first started asking some of its Android users if they wanted to open a link before retweeting it. Twitter now says the prompt has been a success. The experiment has led to more people opening articles, and in some cases opting not to send the retweet after all. 

The Verge / Twitter

The Verge also got to see the interface for voice messages. At the moment, it’s a simple affair. Playback control is limited to playing and pausing a message, but you can record a missive directly in-line. Rounding things out is an option to report a message if someone is misusing the feature.  

While audio messages are relatively new ground for Twitter, it’s a feature you can find in many other apps. Facebook, for instance, allows you to send them in Messenger, Instagram and WhatsApp. As Twitter pointed out when it started testing audio tweets on iOS, voice messages add a “more human touch” to the way people use a chat platform. So it’s not a surprise to see the company willing to expirment with the format for DMs.

As part of National Voter Registration Day, Twitter is rolling out new tools to connect US users with voter registration resources. The platform says this is “its biggest push ever” to encourage people to register.

Today, every person with a Twitter account in the US will see a timeline prompt encouraging them to register to vote or confirm their registration with TurboVote. A push alert (available in over 40 languages) will direct app users to additional voter registration information, and users will see #NationalVoterRegistrationDay and #VoteReady.

Twitter is learning first-hand about the challenges of eliminating racial bias in algorithms. The social network’s Liz Kelley said the company had “more analysis” to do after cryptographic engineer Tony Arcieri conducted an experiment suggesting Twitter’s algorithm was biased in prioritizing photos. When attaching photos of Barack Obama and Mitch McConnell to tweets, Twitter seemed to exclusively highlight McConnell’s face — Obama only popped up when Arcieri inverted the colors, making skin color a non-issue.

Others tried reversing photo and name orders to no avail. A higher-contrast smile did work, Intertheory’s Kim Sherrell found. Scientist Matt Blaze, meanwhile, noticed that the priority seemed to vary depending on the official Twitter app used. Tweetdeck was more neutral, for instance.

In addition to the new login and password protection measures, Twitter says it will take additional steps to protect accounts internally. These include:

-More sophisticated detections and alerts to help us, and account holders, respond rapidly to suspicious activity

-Increased login defenses to prevent malicious account takeover attempts

-Expedited account recovery support to ensure account security issues are resolved quickly

The new measures will apply to politicians at the federal and state level, as well as people associated with political parties, campaigns, and candidates in US House, Senate or gubernatorial races. The requirements will also apply to “major US news outlets and political journalists.” 

Lawmakers have previously criticized Twitter for not doing enough to protect high-profile accounts after hackers gained control of several prominent accounts, including Barack Obama, Joe Biden, Elon Musk and Bill Gates. A Florida teenager and two others were later arrested for their alleged roles in the hack.

The update is the latest in a series of changes Twitter has implemented to strengthen its policies and security ahead of the election. The company also introduced a new misinformation policy and added a hub for election news to promote information from “reputable” sources. 

Twitter has banned the accounts of far-right extremist group the Oath Keepers and its founder, Stewart Rhodes. According to the Southern Poverty Law Center, the group is one of the US’s largest anti-government movements. It claims to have tens of thousands of members, many of whom identify as former law enforcement and military. The group recently tweeted there would be “open warfare against the Marxist insurrectionists (its shorthand for Black Lives Matter activists) by election night.”

A spokesperson for Twitter told Engadget it banned accounts associated with the Oath Keepers for violating its policies on violent extremist groups. “There is no place on Twitter for terrorist organizations or violent extremist groups and individuals who affiliate with and promote their illicit activities,” the company says on the policy page. “We examine a group’s activities both on and off Twitter to determine whether they engage in and/or promote violence against civilians to advance a political, religious and/or social cause.”

Notably, this includes not just claims about “election rigging,” and “claiming victory before election results have been certified,” but tweets “inciting unlawful conduct to prevent a peaceful transfer of power or orderly succession.”

Here’s Twitter’s full explanation of the type of claims it will be looking out for:

False or misleading information that causes confusion about the laws and regulations of a civic process, or officials and institutions executing those civic processes.

Disputed claims that could undermine faith in the process itself, e.g. unverified information about election rigging, ballot tampering, vote tallying, or certification of election results.

Misleading claims about the results or outcome of a civic process which calls for or could lead to interference with the implementation of the results of the process, e.g. claiming victory before election results have been certified, inciting unlawful conduct to prevent a peaceful transfer of power or orderly succession.

The new rules come amid growing pressure on social media companies to do more to fight misinformation ahead of the 2020 presidential election. With the coronavirus pandemic still not under control, record numbers of mail-in ballots are expected. And Donald Trump has already tested Twitter’s existing rules with repeated, and baseless, claims that voting by mail will lead to widespread fraud. 

Experts have warned of various “nightmare scenarios” that could play out in the days and weeks after the election should Trump or others use social media to fuel unverified claims of a “rigged” election. 

Twitter isn’t the only company trying to shore up its rules ahead of the election. Facebook has also updated its policies around misinformation, and both companies recently participated in a meeting with government agencies to discuss election security issues. 

Twitter is launching two new teams: the Accessibility Center of Excellence (ACE) and the Experience Accessibility Team (EAT, that can’t be intentional). While those names sound a bit like they came out of a corporate jargon generator, the teams will do important work. ACE will make aspects of Twitter — from office spaces to marketing and legal and policy standards — more accessible. EAT will work to make new and existing products and features more accessible. Already, Twitter is working to add automated captions to audio and video by early 2021.

It’s great that Twitter listened to feedback and is launching these teams, but it makes you wonder why the company didn’t have such teams in place already. Accessibility concerns are nothing new, and other companies already have teams like this. 

The Americans With Disabilities Act was signed into law 30 years ago. Federal law requires accessibility from the start. You don’t, as a matter of civil rights law, get to roll out an inaccessible feature and then, only later, make it accessible. https://t.co/f58t9pthAy

— Matthew Cortland, Esq (@mattbc) June 17, 2020