Twitter

Bot Tweeted Names And Photos Of Venmo Users Who Bought Drugs (mercurynews.com) 86

Since Venmo's transactions are "public" by default and broadcast on Venmo's API, a Python programmer decided to publicize a few of them, reports the Mercury News: The creator of the bot named "Who's buying drugs on Venmo" under the Twitter handle @venmodrugs says he wanted users to consider their privacy settings before using Venmo. The bot finds Venmo transactions that include words such as heroin, marijuana, cocaine, meth, speed or emojis that denote drugs and tweets the transaction with the names of the sender and receiver and the sender's photo, if there is one... "I wanted to demonstrate how much data Venmo was making publicly available with their open API and their public by default settings and encourage people to consider their privacy settings," Joel Guerra, the creator of the bot, told Motherboard, a technology news outlet run by Vice.
He shut the bot after 24 hours, according to a Medium essay titled "Why I blasted your 'drug' deals on Twitter": I chose drugs, sex and alcohol keywords as the trigger for the bot because because they were funny and shocking. I removed the last names of users because I didn't want to actually contribute to the problem of lack of privacy... I braced myself for backlash but the response was overwhelmingly positive. People understood my point and I had sparked a lot of discussion about online privacy and the need for users to do a better job of understanding the terms of software they were using -- and a lot of discussion about how companies need to do a better job of informing customers how their data was being used...

After about 24 hours of tweeting everyone's drug laden Venmo transactions I shut down the bot (Python script!!) and deleted all the tweets. I had successfully made my point and gotten more attention than I had imagined possible. Thousands of people were reading tweets and articles about the bot and discussing data privacy. I saw no further value in tweeting out anyone's personal transactions anymore. However, all I ever did was format the data and automate a Twitter account -- the data is still readily available.

His closure of the bot drew some interesting reactions on Twitter.

"booooooooo. I was so entertained by this."

"I remember I had a dealer take my phone and set venmo to private lol."

"we're looking to add a Python developer to our team and I think you'd be a good fit."
Privacy

Uber Bans Driver Who Secretly Livestreamed Hundreds of Passengers (mashable.com) 116

Lauren Weinstein tipped us off to this story from Mashable: Hundreds of Uber and Lyft rides have been broadcast live on Twitch by driver Jason Gargac this year, St. Louis Post-Dispatch reported Saturday, all of them without the passengers' permission. Gargac, who goes by the name JustSmurf on Twitch, regularly records the interior of his car while working for Uber and Lyft with a camera in the front of the car, allowing viewers to see the faces of his passengers, illuminated by his (usually) purple lights, and hear everything they say. At no point does Gargac make passengers aware that they are being filmed or livestreamed.

Due to Missouri's "one-party consent" law, in which only one party needs to agree to be recorded for it to be legal (in this case, Gargac is the consenting one), what Gargac is doing is perfectly legal. That doesn't mean it's not 100 percent creepy. Sometimes, to confirm who they are for their driver, the passengers say their full names. Not only that, Gargac has another video that shows the view out the front of his car so that people can see where he's driving, giving away the locations of some passengers' homes.

All the while, viewers on Twitch are commenting about things like the quality of neighborhoods, what the passengers are talking about, and of course, women's looks. Gargac himself is openly judgmental about the women he picks up, commenting to his viewers about their appearances before they get in his car and making remarks after he drops them off. He also regularly talks about wanting to get more "content," meaning interesting people, and is open about the fact that he doesn't want passengers to know they are on camera.

"I feel violated. I'm embarrassed," one passenger told the St. Louis Post-Dispatch. "We got in an Uber at 2 a.m. to be safe, and then I find out that because of that, everything I said in that car is online and people are watching me. It makes me sick."

The offending driver announced today on Twitter that he's at least "getting rid of the stored vids." He calls this move "step #1 of trying to calm everyone down." Hours ago his Twitch feed was made inaccessible.

Lyft and Twitch have not yet responded to Mashable's request for a comment. But Uber said they've (temporarily?) banned Gargac from accessing their app "while we evaluate his partnership with Uber."
Censorship

Lawmakers Call On Amazon and Google To Reconsider Ban On Domain Fronting (cyberscoop.com) 44

An anonymous reader quotes CyberScoop: Amazon and Google face sharp questions from a bipartisan pair of U.S. senators over the tech giants' decisions to ban domain fronting, a technique used to circumvent censorship and surveillance around the world. Sen. Ron Wyden, D-Ore., and Sen. Marco Rubio, R-Fla., sent a letter on Tuesday to Google CEO Larry Page and Amazon CEO Jeff Bezos over decisions by both companies in April to ban domain fronting.

Amazon then warned the developers of encrypted messaging app Signal that the organization would be banned from Amazon's cloud services if the service didn't stop using Amazon's cloud as cover. "We respectfully urge you to reconsider your decision to prohibit domain fronting given the harm it will do to global internet freedom and the risk it will impose upon human rights activists, journalists, and others who rely on the internet freedom tools," the senators wrote.

Robotics

State Senator Wants A Law Forcing Bots To Admit They're Not Human (brisbanetimes.com.au) 151

An anonymous reader writes: Several commentators are calling for a law that requires bots to admit they are not human. There is a bill in California that would do just that. A new paper argues that these laws may look Constitutional but actually raise First Amendment issues.
The New York Times reports: Bots are easy to make and widely employed, and social media companies are under no legal obligation to get rid of them. A law that discourages their use could help, but experts aren't sure how the one [state senator Robert] Hertzberg is trying to push through, in California, might work. For starters, would bots be forced to identify themselves in every Facebook post? In their Instagram bios? In their Twitter handles? The measure, SB-1001, a version of which has already left the senate floor and is working its way through the state's Assembly, also doesn't mandate that tech companies enforce the regulation. And it's unclear how a bill that is specific only to California would apply to a global internet...

All parties agree that the bill illustrates the difficulty that lawmakers have in crafting legislation that effectively addresses the problems constituents confront online. As the pace of technological development has raced ahead of government, the laws that exist on the books -- not to mention some lawmakers' understandings of technology -- have remained comparatively stagnant.

The Times seems to question whether the law should be targeted at the creators of bots instead of the platforms that host them, pointing out that tech companies like Twitter "have the power to change dynamics on their platforms directly and at the scale that those problems require."
Social Networks

Social Media Manipulation Rising Globally, New Oxford Report Warns (phys.org) 99

A new report from Oxford University found that manipulation of public opinion over social media platforms is growing at a large scale, despite efforts to combat it. "Around the world, government agencies and political parties are exploiting social media platforms to spread junk news and disinformation, exercise censorship and control, and undermine trust in media, public institutions and science," reports Phys.Org. From the report: "The number of countries where formally organized social media manipulation occurs has greatly increased, from 28 to 48 countries globally," says Samantha Bradshaw, co-author of the report. "The majority of growth comes from political parties who spread disinformation and junk news around election periods. There are more political parties learning from the strategies deployed during Brexit and the U.S. 2016 Presidential election: more campaigns are using bots, junk news, and disinformation to polarize and manipulate voters."

This is despite efforts by governments in many democracies introducing new legislation designed to combat fake news on the internet. "The problem with this is that these 'task forces' to combat fake news are being used as a new tool to legitimize censorship in authoritarian regimes," says Professor Phil Howard, co-author and lead researcher on the OII's Computational Propaganda project. "At best, these types of task forces are creating counter-narratives and building tools for citizen awareness and fact-checking." Another challenge is the evolution of the mediums individuals use to share news and information. "There is evidence that disinformation campaigns are moving on to chat applications and alternative platforms," says Bradshaw. "This is becoming increasingly common in the Global South, where large public groups on chat applications are more popular."

Government

FCC Opens Public Comments On T-Mobile-Sprint Merger (engadget.com) 43

Now is your chance to voice your opinion on the $26 billion merger of T-Mobile and Sprint. The FCC is now accepting comments as well as formal petitions to deny the merger until August 27th. The companies and supporters of the deal can then file oppositions to those petitions by September 17th, while a final round of replies has a deadline of October 9th. Engadget reports: Anyone can file petitions to deny, and you might expect to see some from consumer advocacy groups and industry experts who may be concerned over the reduction in the number of national carriers from four to three. The FCC has laid out a 180-day review timeline to determine whether the merger is in the public interest, but that's more of a guideline and there's no required deadline for the agency to issue a decision.
Youtube

PeerTube, the 'Decentralized YouTube,' Succeeds In Crowdfunding (quariety.com) 100

A crowdfunded project, known as "PeerTube," has blown through its initial goal with 53,100 euros collected in forty-two days. The project aims to be "a fully decentralized version of YouTube, whose computer code is freely accessible and editable, and where videos are shared between users without relying on a central system." The goal is PeerTube to officially launch by October. Quariety reports: PeerTube relies on a decentralized and federative system. In other words, there is no higher authority that manages, broadcasts and moderates the content offered, as is the case with YouTube, but a network of "instances." Created by one or more administrators, these communities are governed according to principles specific to each of them. Anyone can freely watch the videos without registering, but to upload a video, you must choose from the list of existing instances, or create your own if you have the necessary technical knowledge. At the moment, 141 instances are proposed. Most do not have specifics, but one can find communities centered on a theme or open to a particular region of the world. In all, more than 4,000 people are currently registered on PeerTube, for a total of 338,000 views for 11,000 videos. The project does not display ads, unlike YouTube. "In terms of monetization, we wanted to make a neutral tool," says Pouhiou, communication officer for Framasoft, the origin of PeerTube. The site will rely on a "support" button at the start, but "people will be able to code their own monetization system" in the future.
Businesses

Uber Drivers 'Employees' For Unemployment Purposes, New York Labor Board Says (arstechnica.com) 54

An anonymous reader quotes a report from Ars Technica: New York City's largest taxi driver advocacy group is hailing a legal decision by the New York State Unemployment Insurance Appeal Board, which ruled last Friday that three out-of-work Uber drivers can be considered employees for the purpose of unemployment benefits. The decision was first reported Thursday by Politico. In other words, three men -- and possibly other "similarly situated" Uber drivers who had quit over low pay or who were deactivated from the Uber platform -- can get paid. "The decision means that New York Uber drivers can file for unemployment insurance and likely receive it," Veena Dubal, a labor law professor at the University of California Hastings College of the Law in San Francisco, emailed Ars. "Uber may appeal the decision to state court, but for now, it's good law."
Privacy

Venmo Refuses To Say Why Transactions Are Public By Default (zdnet.com) 47

Venmo, the mobile payments app, won't say why it exposes users' data to the world whenever they make a transaction. ZDNet: Hang Do Thi Duc, a Berlin-based privacy researcher found that every time someone sent or received money using the PayPal-owned mobile app (which had over seven million users in 2017), the transaction was "public" by default and was broadcast on Venmo's API. In other words, everyone can see your transactions -- even without the app. The company did not respond to ZDNet's queries, but in a blanket statement said it takes privacy of users seriously.

Further reading: People Are Using Venmo To Spy On Cheating Spouses.
Security

Chinese Hackers Targeted IoT During Trump-Putin Summit (defenseone.com) 82

Zorro shares a report from Defense One: Four days before U.S. and Russian leaders met in Helsinki, hackers from China launched a wave of brute-force attacks on internet-connected devices in Finland, seeking to gain control of gear that could collect audio or visual intelligence, a new report says. Traffic aimed at remote command-and-control features for Finnish internet-connected devices began to spike July 12, according to a July 19 report by Seattle-based cybersecurity company F5.

China generally originates the largest chunk of such attacks; in May, Chinese attacks accounted for 29 percent of the total. But as attacks began to spike on July 12, China's share rose to 34 percent, the report said. Attacks jumped 2,800 percent. The China-based hackers' primary target was SSH (or Secure Shell) Port 22 -- not a physical destination but a specific set of instructions for routing a message to the right destination when the message hits the server. "SSH brute force attacks are commonly used to exploit systems and [internet of things, or IOT] devices online," the report says. "SSH is often used by IoT devices for 'secure' remote administration."
The report notes that attack traffic came from the U.S., France, and Italy as well, but the U.S. and French traffic kept with its averages. "Russian attack traffic dropped considerably from third, its usual spot, to fifth," reports Defense One. "German attack traffic jumped."
Businesses

FCC Vote Likely Dooms Sinclair-Tribune Merger (engadget.com) 37

FCC commissioners unanimously voted on a Hearing Designation Order (HDO) to send the proposed sale of Tribune Media properties to Sinclair to a judge, where the merger is expected to cease. Engadget reports: Earlier this week, FCC chairman Ajit Pai raised "serious concerns" about Sinclair's selloff of 21 stations it had proposed in order to remain under station ownership limits post-merger. Had Sinclair declined to sell off some stations, its 173 broadcast stations in 81 markets, combined with Tribune's 42 stations in 33 markets would reach 72 percent of U.S. TV households. The FCC's National TV Ownership rule "does not limit the number of TV stations a single entity may own nationwide so long as the station group collectively reaches no more than 39 percent of all U.S. TV households." But the rule is more flexible for stations that broadcast using UHF frequencies. Pai, who has been accused of aiding the merger by relaxing the ownership regulations, said Monday that Sinclair's plan would allow the company "to control those stations in practice, even if not in name, in violation of the law." He noted that, "When the FCC confronts disputed issues like these, the Communications Act does not allow it to approve a transaction."
Government

FBI Director: Without Compromise on Encryption, Legislation May Be the 'Remedy' (cyberscoop.com) 393

An anonymous reader shares a report: FBI Director Christopher Wray said Wednesday that unless the U.S. government and private industry are able to come to a compromise on the issue of default encryption on consumer devices, legislation may be how the debate is ultimately decided. "I think there should be [room for compromise]," Wray said Wednesday night at a national security conference in Aspen, Colorado. "I don't want to characterize private conversations we're having with people in the industry. We're not there yet for sure. And if we can't get there, there may be other remedies, like legislation, that would have to come to bear." Wray described the issue of "Going Dark" because of encryption as a "significant" and "growing" problem for federal, state and local law enforcement as well as foreign law enforcement and intelligence agencies. He claims strong encryption on mobile phones keeps law enforcement from gaining access to key evidence as it relates to active criminal investigations. "People are less safe as a result of it," he said.
Communications

Senate Wants Netflix, Spotify To Send Out Federal Emergency Alerts (techcrunch.com) 107

Senators in Hawaii and South Dakota have introduced a bill, called the "Reliable Emergency Alert Distribution Improvement (READI) act, that would "explore" broadcasting alerts to "online streaming services, such as Netflix and Spotify," amongst other changes to the Emergency Alert System. TechCrunch reports: Some of the other things the bill touches on:
- Users on many phones can currently disable federal alerts; they want to get rid of that option
- Building a better system for reporting false alarms and figuring out what happened
- Updating the system to better prevent false alarms, and to better retract them when they do happen

Security

Microsoft Reveals First Known Midterm Campaign Hacking Attempts (politico.com) 163

An anonymous reader shares a report: Microsoft detected and helped block hacking attempts against three congressional candidates this year, a company executive said Thursday, marking the first known example of cyber interference in the midterm elections. "Earlier this year, we did discover that a fake Microsoft domain had been established as the landing page for phishing attacks," said Tom Burt, Microsoft's vice president for security and trust, at the Aspen Security Forum. "And we saw metadata that suggested those phishing attacks were being directed at three candidates who are all standing for election in the midterm elections."

Burt declined to name the targets but said they were "people who, because of their positions, might have been interesting targets from an espionage standpoint as well as an election disruption standpoint." Microsoft took down the fake domain and worked with the federal government to block the phishing messages.

Government

Should the Word 'Milk' Be Used To Describe Nondairy Milk-Alternative Products? (arstechnica.com) 520

An anonymous reader quotes a report from Ars Technica: The U.S. Food and Drug Administration seems to have soured on nondairy milk-alternative products that use the term "milk" in their marketing and labeling -- like popular soy and almond milk products. In a talk hosted by Politico, FDA Commissioner Scott Gottlieb announced Tuesday that the FDA will soon issue a new guidance on the use of the term. But he added that products aren't abiding by FDA policies as they stand now. He referenced a so-called "standard of identity" policy that regulates how milk is defined and should be identified. "If you look at our standard of identity -- there is a reference somewhere in the standard of identity to a lactating animal," he said. "And, you know, an almond doesn't lactate, I will confess."

He went on to explain that the issue is that the agency hasn't been enforcing its own policy or putting the squeeze on product makers -- and that it's time to get abreast of the labeling language. But, he admitted, curtailing the wording of non-moo juice labeling isn't an easy task because it means that the agency has to change its "regulatory posture." "I can't just do it unilaterally," Gottlieb said. Hence, the agency is putting together a new guidance for manufacturers to help skim the fat from the market. Gottlieb said the agency will soon tap the public for comments on the terminology and hopes to wring out a new policy within a year.

Earth

Russian Shipwreck Allegedly Carrying $130 Billion In Gold Has Been Rediscovered (popularmechanics.com) 256

An anonymous reader quotes a report from Popular Mechanics: A salvage company has located the remains of a Russian warship lost during the the Russo-Japanese War. The battle-damaged cruiser Dmitrii Donskoi was scuttled off the coast of Korea in 1905, reportedly carrying a cargo of gold worth an estimated $130 billion in today's dollars. An international consortium of companies plans to salvage the gold.

According to the Telegraph, the Donskoi was found less than a mile off the coast of Ulleung island, at a depth of 1,423 feet in the Sea of Japan. A submersible descended to the wreck and captured an image of the ship's name on the stern in the Cyrillic alphabet. The South Korean Shinil Group, which discovered the wreck, plans to recover the gold sometime later this year with help from companies in China, Canada, and the U.K. At the time of her sinking Donskoi was reportedly carrying 5,500 boxes of gold bars and 200 tons of gold coins with a street value today of $130 billion. That's more than twice Russia's 2017 defense budget, which was $61 billion. If the treasure does materialize, the Russian government will receive half of the recovered amount.
The money that's not going to Russia will reportedly be invested in a railroad line linking North Korea, South Korea, and Russia. A small percentage (10%) will also be invested in tourism projects on Ulleungdo Island, including a museum dedicated to the vessel.
Communications

Leaked Documents Show Facebook's 'Threshold' For Deleting Pages, Groups (vice.com) 94

Facebook has repeatedly referenced to lawmakers a "threshold" that must be reached before the platform decides to ban a particular page for violating the site's policies, but it hasn't discussed its guidelines publicly. Motherboard has obtained internal Facebook documents laying out what this threshold is for multiple types of different content, including some instances of hate speech. From the report: One Facebook moderator training document for hate speech says that for Pages -- Facebook's feature for sections dedicated to, say, a band, organization, public figure, or business -- the Page admin has to receive 5 "strikes" within 90 days for the Page itself to be deleted. Alternatively, Facebook moderators are told to remove a Page if at least 30 percent of the content posted by other people within 90 days violates Facebook's community standards. A similar 30 percent-or-over policy exists for Facebook Groups, according to the document.

In a similar vein, another hate speech document says that a profile should be taken down if there are 5 or more pieces of content from the user which indicate hate propaganda, photos of the user present with another identifiable leader, or other related violations. Although the documents obtained by Motherboard were created recently, Facebook's policies change regularly, so whether these exact parameters remain in force is unclear. Of course this still depends on moderators identifying and labeling posts as violating to reach that threshold. [...] Another document focused on sexual content says moderators should unpublish Pages and Groups under the basis of sexual solicitation if there are over 2 "elements," such as the Page description, title, photo, or pinned post, that include either explicit solicitation of nude imagery, or, if the page is more subtle, includes either a method of contact or a location. This slide again reiterates the over 30 percent and 5 admin posts rules found in the hate speech document.

AI

DeepMind, Elon Musk and Others Pledge Not To Make Autonomous AI Weapons (engadget.com) 122

An anonymous reader quotes a report from Engadget: Yesterday, during the Joint Conference on Artificial Intelligence, the Future of Life Institute announced that more than 2,400 individuals and 160 companies and organizations have signed a pledge, declaring that they will "neither participate in nor support the development, manufacture, trade or use of lethal autonomous weapons." The signatories, representing 90 countries, also call on governments to pass laws against such weapons. Google DeepMind and the Xprize Foundation are among the groups who've signed on while Elon Musk and DeepMind co-founders Demis Hassabis, Shane Legg and Mustafa Suleyman have made the pledge as well.

"Thousands of AI researchers agree that by removing the risk, attributability and difficulty of taking human lives, lethal autonomous weapons could become powerful instruments of violence and oppression, especially when linked to surveillance and data systems," says the pledge. It adds that those who sign agree that "the decision to take a human life should never be delegated to a machine."
"I'm excited to see AI leaders shifting from talk to action, implementing a policy that politicians have thus far failed to put into effect," Future of Life Institute President Max Tegmark said in a statement. "AI has huge potential to help the world -- if we stigmatize and prevent its abuse. AI weapons that autonomously decide to kill people are as disgusting and destabilizing as bioweapons, and should be dealt with in the same way."
IBM

IBM Wants $167 Million From Groupon Over Alleged Patent Infringement (reuters.com) 64

On Monday, IBM asked a jury to award the company $167 million in a lawsuit against deals site Groupon for using patented technology without authorization. The patents involve e-commerce technology that had already been licensed to Amazon, Facebook, and Alphabet for between $20 million and $50 million per company. "Most big companies have taken licenses to these patents," IBM's lawyer, John Desmarais, said. "Groupon has not. The new kid on the block refuses to take responsibility for using these inventions." Reuters reports: Groupon lawyer J. David Hadden argued that IBM was overreading the scope of its patents and claiming ownership of building blocks of the internet. "A key question for you in this case is whether these patents cover the world wide web," Hadden told jurors. "They do not and that is because IBM did not invent the world wide web."

An IBM executive is expected to testify during the two-week trial about licensing deals with technology companies like Amazon and Google, providing a rare glimpse into IBM's efforts to derive revenue from its large patent portfolio. The Armonk, New York-based company invests heavily in research and development and has secured more U.S. patents than any other company for the past 25 years.

Security

Hackers Account For 90 Percent of Login Attempts At Online Retailers (qz.com) 33

Hackers account for 90% of of e-commerce sites' global login traffic, according to a report by cyber security firm Shape Security. They reportedly use programs to apply stolen data acquired on the dark web -- all in an effort to login to websites and grab something of value like cash, airline points, or merchandise. Quartz reports: These attacks are successful as often as 3% of the time, and the costs quickly add up for businesses, Shape says. This type of fraud costs the e-commerce sector about $6 billion a year, while the consumer banking industry loses out on about $1.7 billion annually. The hotel and airline businesses are also major targets -- the theft of loyalty points is a thing -- costing a combined $700 million every year.

The process starts when hackers break into databases and steal login information. Some of the best known "data spills" took place at Equifax and Yahoo, but they happen fairly regularly -- there were 51 reported breaches last year, compromising 2.3 billion credentials, according to Shape. Taking over bank accounts is one way to monetize stolen login information -- in the US, community banks are attacked far more than any other industry group. According to Shape's data, that sector is attacked more than 200 million times each day.
Shape says the number of reported credential breaches was roughly stable at 51 last year, compared with 52 in 2016. The best way consumers can minimize these attacks is by changing their passwords.

Slashdot Top Deals