Tag Archives: censorship

Under pressure, Twitter prepares to extend reporting abuse to all users

Under increased scrutiny, Twitter will be extending the ability to report tweets to all of its hundreds of millions of active users around the world.

A statement from Twitter, emailed to the BBC and GigaOm, urged users to report abusive behavior and violations of the relevant policy and Twitter Rules using an online form and shared plans to “bring the functionality to other platforms, including Android and the web.” Twitter hasn’t shared timelines for that extension yet, but aggrieved users in Britain and beyond should gain the ability to flag tweets with a couple of taps eventually.

report-tweet-button

Twitter users have been able to report violations and abuse for years, with decisions by the service’s Safety team as tickets or law enforcement interest comes in. Twitter’s Safety team, headed by Del (@delbius) Harvey, has been quietly, professionally handling the ugly side for many years.

Adding reporting to individual tweets, however, is a relatively new change that was not announced on the Twitter blog or through the @Safety or @Support accounts.

Here are the relevant details from Twitter’s FAQ:

You can report Tweets that are in violation of the Twitter Rules or our Terms of Service. This includes spam, harassment, impersonation, copyright, or trademark violations. You can report any Tweet on Twitter, including Tweets in your home timeline, the Connect or Discover tabs, or in Twitter Search.

To report a Tweet:

  • Navigate to the Tweet you’d like to report.
  • Tap the ••• icon to bring up the off-screen menu.
  • Select Report Tweet and then one of the options below.
  • Select Submit (or Next if reporting abuse; see below for details) or Cancel to complete the report or block the user.

Reporting options:

Spam: this is the best option for reporting users who are using spam tactics. Please reference the Twitter Rules for information about some common spam techniques, which include mass creation of accounts for abusive purposes, following a large number of users in a short time, and sending large numbers of unsolicited @replies.
Compromised: if you think the user’s account has been compromised, and they are no longer in control of their account, select this option, and we will follow up with them to reset their password and/or take other appropriate actions.
Abusive: for other types of violations, including harassment, copyright or trademark violations, and impersonation, select this option. When you select “Next’”, you’ll be taken to a form where you can complete and submit your report to Twitter.
Block account: instead of reporting a user, you can select this option to block the user. If you block a user, they will not be allowed to follow you or add you to lists, and you won’t see any interactions with the user in your Connect tab.

Icebergs ahead

Twitter has successfully scaled the ability to flag media to all of its users. They’ve kept the Fail Whale from surfacing by vastly increasing the capacity of the service to handle billions of tweets and surges in use during major events. They’ve already rolled out tweet reporting to Twitter to iPhone users. Now, they’ll simplify reporting of abuse tweets for everyone.

There may be hidden tradeoffs in adding this function, as Staci Kramer pointed out on Twitter: previously available options, like “tweet link,” “mail link” and “read later” aren’t in the new version of Twitter’s iOS app.

What may prove more difficult than adding this function to other official apps and the Web, however, will be adding the human capacity to adjudicate decisions to suspend or restore accounts.

Twitter will be doing it under increasing scrutiny and a fresh wave of critics who are taking the company to task for being slow to respond to threats and abuse. More than 18,000 people have signed a petition at Change.org demanding that Twitter provide a an abuse reporting button. The petition was filed after a stream of rape threats were directed at Caroline Criado-Perez on Twitter for 48 hours.

Criado-Perez, a freelance journalist and self-described feminist campaigner, was in the public eye because of her successful efforts to keep pictures of women on paper money. She began receiving abusive tweets on the day that the Bank of England announced that author Jane Austen would appear on its newly designed £10 note.

The signatories on the petition were asking for a function that already exists for the millions of Twitter users that access the service on an iPhone, as the head of the social networking service’s United Kingdom office tweeted earlier today, responding to heated criticism in the British press.

To mollify critics and offer a users a better experience, Twitter staff will need to proactively detect waves of abuse, aided by algorithms and adjudication systems, and make judgements about whether tweets break its stated policies or represent threats that must be reported to law enforcement.

“I don’t know what proportion of posts are abusive, nor do I know the volume of complaints handled by Twitter staff and their response time, which are obvious factors in how and when abuse reports are handled,” commented veteran journalist Saleem Khan. “If there’s a problem with complaint-handling, Twitter needs to examine its processes and staffing. That said, if abuse and/or non-responsiveness by staff are perceived to be a problem, then it is a problem.”

To state the obvious, this will be an ongoing headache for Twitter.

Like other social media companies, Twitter has been navigating deep, troubled currents of censorship, privacy and suspensions in recent years.

Creating systems that offer fair, efficient moderation and adjudication of reports is a conundrum that code alone may not be able to solve. That challenge is extended by the presence of organized campaigns of humans and bots that game governance systems by flagging users en masse as spammers, leading to suspensions.

That may well mean that Twitter, like other social networks with millions of users, will need to expand its safety team and train the rest of its public-facing employees to act as ad hoc ombudsmen and women, as aggrieved users inevitably turn their ire upon staff using the network. They’re well positioned to do so, perhaps better than any other social network, but the service is inevitably going to face tough decisions as it operates in countries do not have legal protections for freedom of expression or the press.

As Rebecca MacKinnon, Ethan Zuckerman and others have highlighted, what we think of as the new public square online is owned and operated by private companies that are setting the terms and conditions for expression and behavior on them. Giving users the capacity to report abuse, fraud or copyright infringement is a natural feature for any major website or service but it comes with new headaches. If Twitter is to go public, however, it will need to develop more matures to handle being a platform for the public.

“The question remains,” commented Khan: “What rights and powers do we delegate to private, for-profit, unregulated platforms that increasingly mediate the majority of our discourse, and where is the line that we draw in that deal?”


Editor’s Note: I sent Twitter a series of questions regarding the new reporting function on Sunday morning. On Sunday night, Twitter declined to comment further than the statement they have released. On Monday afternoon, Twitter CEO Dick Costolo responded to tweeted queries. Following are the questions I posed over email. If you have answers, feel free to comment or contact me.

When was this added? Was there an official blog post or tweets from staff, @safety and @support about it?

What’s the timeline for it rolling out to all users? Will Twitter for Windows and BlackBerry and get it?

Will it be added to the API, so that TweetBot and TweetDeck users, along with other clients, can use it after updates?

Will Twitter increase staffing at Safety and Support to handle an increase in reports? To what levels?

Will there be designated ombudsmen or women?

Will there be any transparency into the number of tickets received regarding abuse or someone’s status in the queue?

Will Twitter release aggregate data of abuse (or spam) flagging? How will Twitter deal with false positives or organized/automated campaigns to flag users or tweets?

Will there be any consequences for users that repeatedly abuse the ability to flag people or tweets for abuse?


Postscript

On August 3, Twitter responded with an update to its rules to help address abusive behavior, including extra staff to handle abuse reports.

“It comes down to this: people deserve to feel safe on Twitter,” said Twitter’s UK lead Tony Wang and Del Harvey, senior director for trust and safety, in a blog post.

We want people to feel safe on Twitter, and we want the Twitter Rules to send a clear message to anyone who thought that such behaviour was, or could ever be, acceptable.”

The updated rules apply globally. “As described in the blog post, this was a clarification of existing rules — we discussed harassment in our help center in connection with abuse, but this makes it explicit in the rules as well,” said Twitter communication lead Jim Prosser, reached by email.

Wang also tweeted an apology to the women who have been targeted by abuse on Twitter.

“I personally apologize to the women who have experienced abuse on Twitter and for what they have gone through,” he said. “The abuse they’ve received is simply not acceptable. It’s not acceptable in the real world, and it’s not acceptable on Twitter.”

So yes, there are limits to free speech on Twitter.

What are they? Well, that’s the sticky wicket. The updated rules now include a section that Harvey said already existed. Twitter “actually always had that as a note on our abusive behavior policy page; we just added it directly to the rules,” she tweeted.

Targeted Abuse: You may not engage in targeted abuse or harassment. Some of the factors that we take into account when determining what conduct is considered to be targeted abuse or harassment are:
*if you are sending messages to a user from multiple accounts;
*if the sole purpose of your account is to send abusive messages to others;
*if the reported behavior is one-sided or includes threats

This was “no real addition, just [a] clarification,” tweeted Harvey. “Twitter “just added the explicit callout to our preexisting policy under the abuse & spam section.”

There is no functional difference in how Twitter’s Safety team will now assess abuse reports, she further clarified.

“We’ve been working on making the reporting process easier for users & clarifying our policies.”

Advertisements

5 Comments

Filed under article, journalism, social media, Twitter

On Twitter suspensions, spam, censorship and SOPA

Suspended OwlEarlier this afternoon, David Seaman claimed that Twitter suspended his account for tweeting too much about “Occupy Wall Street … and talking too much about the controversial detainment without trial provisions contained in the FY 2012 National Defense Authorization Act (NDAA).”

His account is now back online. Twitter’s official response to him, according to Seaman, was that his account was ‘caught up in one of spam groups by mistake.

Seaman continued to suggest otherwise and implied that Twitter is banning accounts because of their content.

Speaking only for myself, I believe this was completely unrelated to NDAA or OWS and was instead tied to his behavior using a new account. I think what happened today was an auto-suspension of a new account exhibiting behavior associated, not intentional censorship by Twitter. Jillian C. York, the director for international freedom of expression at the Electronic Frontier Foundation, agreed:

I’m writing without an official statement from Twitter but I’d bet that’s what happened. (If I receive such a statement, I’ll post it here.)

UPDATE: Here are the emails Seaman posted to his post, containing Twitter’s responses. They validates my understanding of Twitter’s anti-spam protocols.

At approximately 7:37pm ET, my Twitter account was restored, and I received the following message from Twitter support: “Hello, Twitter has automated systems that find and remove multiple automated spam accounts in bulk. Unfortunately, it looks like your account got caught up in one of these spam groups by mistake. I’ve restored your account; sorry for the inconvenience. Please note that it may take an hour or so for your follower and following numbers to return to normal.” At 8:29pm ET, a second email from Twitter support was received: “Hello, As a clarification, your account was suspended twice; the initial suspension was due to a number of unsolicited duplicate or near-duplicate messages being sent using the @reply and/or mention feature. These features are intended to make communication between people on Twitter easier. Twitter monitors the use of these features to make sure they are used as intended and not for abuse. Using either feature to post messages to a bunch of users in an unsolicited or egregious manner is considered an abuse of its use, which results in an automated account suspension. However, the second suspension after you navigated the self-unsuspension page was due to a known error we are working to fix; our apologies for the re-suspension. Please let me know if you have any questions.”

As far as I know, Twitter accounts aren’t automatically suspended based upon a journalist writing about a controversial issue. You can read the Twitter FAQ on suspensions for their official position. Suspensions are only supposed to happen when a user breaks the Twitter Rules, not because of what they describe or report on. Again, York:

Suspending accounts on Twitter is precedented behavior. What’s less so is a self-identified journalist making a sweeping claim of censorship like this without confirmation, corroboration or analysis of Twitter’s past practices. My account was suspended 2 years ago when @Twitter swept it up on people tweeting on the #g2s hashtag. It was restored the day after wards, along with other people tweeting from the IP address.

I doubt Seaman’s contention that this suspension was related to content. I think it was a mistaken outcome based upon interactions. New accounts are more likely to be flagged automatically as @spam. What happened wasn’t about any one tweet: it’s came through nine tweets in a row of nearly duplicate content to non-followers from a new account. Specifically, “How #Occupy and the #TeaParty could end their struggle tonight: http://read.bi/vL02ZI #NDAA #SOPA #OWS”

Bottom line: Seaman made a sensational claim that probably shouldn’t have been made without more legwork and a statement from Twitter. He used Business Insider’s platform to bring attention to a mistake. It may have brought Business Insider a lot of traffic today but I think, on balance, that Seaman damaged his credibility today.

That’s unfortunate, given that the episode could have been leveraged to make an important point about how governments might work with private social media platforms to remove content that they do not wish to see published.

On that count, learn more about the Stop Online Piracy Act at Radar.

Update: Conor Adams Stevens picked up the Business Insider post and wrote a largely uncritical op-ed at International Business Times that repeated the claim that “NDAA, SOPA, Occupy Wall Street and Anonymous may be off-limits on Twitter.” (If that were true, I wouldn’t have been able to tweet for quite a few months now.)

Update: Nick Judd picked up the story at techPresident, adding some context to the latest episode of Twitter denying another censorship accusation. Judd observes that Deamon’s post “appears to be flat out wrong”:

Seaman still seems to think that some occult hand is at work against opponents of NDAA, questioning the veracity of Twitter’s response to him. This makes no sense, given that NDAA has generated at least 117,000 tweets in the last seven days. None of those have been swept under the digital rug.

There’s also a conspiracy theory floating around about why Twitter has not listed NDAA as a trending topic. Mat Honan bursts that bubble in a post from last week for Gizmodo, which is actually focused on a hashtag memorializing the late Christopher Hitchens. Its title is succinct: “Shutup, Twitter Isn’t Censoring Your Dumb Trends.”

Image Credit: Steve Garfield

13 Comments

Filed under blogging, journalism, social media, technology, Twitter

CDT, EFF, CEA, PK, Others Criticize ACTA Copyright Treaty Draft Language

On March 22, a collection of tech advocates, non-profits and associations opposed to the proposed Anti-Counterfeiting Trade Agreement (ACTA) wrote a letter (embedded below) to Ron Kirk, head of the Office of the U.S. Trade Representative, criticizing the reported draft language of the ACTA.

The lack of transparency around ACTA copyright treaty negotiations has received increased scrutiny from both the media and anti-censorship advocates as provisions have leaked online. Earlier this month, the European Parliament passed a resolution by a 663-13 vote calling for the European Commission, the European Union’s regulatory arm, to release a public draft of the ACTA agreement. President Obama signaled his support for the ACTA copyright treaty at a conference in DC on March 12th. New Zealand is now pushing for greater ACTA transparency as well.

This letter states that “details of the text of the proposed ACTA, and comments and proposals of national participants have apparently but unofficially been made public”. See, for example, this EU document,  a working document from the EU Secretariat, document[PDF], and document.

This letter states that “this negotiation is not primarily about counterfeiting or piracy; nor is at all about trade law. The public rationale that the treaty would not impinge on domestic law has been placed in doubt — particularly when one considers whose domestic law would be endangered.”

On this count, the letter states that the “text reveals detailed substantive attention to … The extent to which principles of inducement, newly introduced by the U.S. Supreme Court in the Grokster case, are to be accepted as supporting a separate basis for copyright liability or are a gloss on existing principles of contributory and vicarious infringement. This is not yet clear even in the United States.”

Some of this ground was covered by Supreme Court’s in June 26, 2005, MGM v. Grokster.a More detail on that case may be found in  “Supreme Court Rules in MGM v. Grokster” in TechLawJournal.  a

The letter regarding ACTA transparency also states that the “text reveals detailed substantive attention to … How technological measure anti-circumvention provisions are to be interpreted and applied, whether they will apply to access to works, whether they are to be limited to circumventions for infringing purposes, and whether account will be taken of the variations in national law, practice, and context, such as U.S. adherence to fair use and the imposition of levies under other national law.”

The signers of this letter include:

Canadian law professor Michael Geist’s ACTA coverage has been instrumental to providing details to the global community of the treaty. As he wrote yesterday:

The leak of the full consolidated ACTA text will provide anyone interested in the treaty with plenty to work with for the next few weeks.  While several chapters have already been leaked and discussed (see posts on the Internet and Civil Enforcement chapters, the definitional chapter, the institutional arrangements chapter, and international coooperation chapter), the consolidated chapter provides a clear indication of how the negotiations have altered earlier proposals (see this post for links to the early leaks) as well as the first look at several other ACTA elements.

Nate Anderson over at Ars Technica also wrote an update on the current status of the ACTA copyright treaty earlier this week. As Mike Masnick blogged at TechDirt, “EU Negotiators Insist That ACTA Will Move Forward And There’s Nothing To Worry About.” Further, as Masnick points out, ACTA is set to cover intellectual property, not just copyrights or trademarks, referring to a post at KEI which features leaked draft document of ACTA.

Leave a comment

Filed under technology

Google shuts down Google.cn, adds censorship dashboard | #GoogleCn

Last night, Google shut down its China search engine, Google.cn. Visitors to Google.cn are now redirected to Google’s Chinese-language service based in Hong Kong, Google.com.hk.

Google has now set up a censorship dashboard for Google services in China that shows which services are blocked.

As Ron Deibert of CitizenLab tweeted, “It’s no ONI report, nor Herdict, but interesting anyway.”

In a statement posted to Google’s official blog, David Drummond explained the new approach to China. Google had previously announced on January 12 that it would no longer stand by a 2006 deal with the Chinese government after it was the target of hacker attacks that it attributed to China.

“CDT applauds Google for following through on its commitment to protect human rights and for its continued effort to enable China’s people with unfiltered access to robust sources of information from all over the world,” said Leslie Harris, President and CEO of the Center for Democracy & Technolog.

“Whether the Chinese people will be able to take advantage of Google search now rests squarely with the Chinese government. If China allows access to unfiltered search, it will be a substantial win for global Internet freedom and for the Chinese people. If China blocks access, it will finally make clear to the Chinese people who is pulling the levers of censorship in the country.”

“It is certainly a historic moment,” said Xiao Qiang, director of the China Internet project at the University of California, Berkeley, quoted in “Google Shuts China Site in Dispute Over Censorship,” in the New York Times. “The Internet was seen as a catalyst for China being more integrated into the world. The fact that Google cannot exist in China, clearly indicates that China’s path as a rising power is going in a direction different from what the world expected and what many Chinese were hoping for.”

As the Ryan Singel reports in his post on Epicenter blog at Wired, “Google Uncensors Chinese Search Engine,” “now a search on June 4, the day of the 1989 Tiananmen massacre, returns 226 million results. Formerly that search, and thousands of other terms like it, had limited results and a notification to users that search results had been hidden due to the rules of China’s Communist government.”

Now, Chinese Internet users are braced to lose Google, as Kathryn Hille reports in the Financial Times.  Bobbie Johnson is liveblogging further developments and statements regarding the shutdown of Google’s search engine in China at the Guardian.

Rebecca McKinnon is also tweeting news and reactions from China. MacKinnon’s interview with Google’s David Drummond on Google and China is a must-read.

UPDATE: Danny Sullivan has also weighed in: “Google Stops Censoring In China, Hopes Using New Domain Meets Legal Requirements.”

Leave a comment

Filed under blogging, technology