The question comes up over and over, with extremist material, hate speech, election meddling and privacy invasions. Why can’t Facebook just fix it?

It’s complicated, with reasons that include Facebook’s size, its business model and technical limitations, not to mention years of unchecked growth. Oh, and the element of human nature.

The latest revelation: Facebook is inadvertently creating celebratory videos using extremist content and auto-generating business pages for the likes of Islamic State and al-Qaida. The company says it is working on solutions and the problems are getting better. That is true, but critics say better is not good enough when mass shootings are being live-streamed and online mobs are spreading rumors that lead to deadly violence.

“They have been frustratingly slow in dealing with everything from child sexual abuse to terrorism, white supremacy, bullying, nonconsensual porn” and things like allowing advertisers to target categories such as “Jew hater,” simply because some users had listed the term as an “interest,” said Hany Farid, a digital forensics expert at the University of California, Berkeley.

As new problems crop up, Facebook’s formula has been to apologize and promise to make changes, sometimes also noting that it did not anticipate how malicious actors could so readily misuse its platform. More recently, the company has also emphasized just how much it is improving, both technically in its use of artificial intelligence to detect problems and in terms of focusing more money and effort on fixing them.

“After making heavy investments, we are detecting and removing terrorism content at a far higher success rate than even two years go,” Facebook said Wednesday in response to the revelations about the auto-generated pages. “We don’t claim to find everything, and we remain vigilant in our efforts against terrorist groups around the world.”

It has seen some success. In late 2016, CEO Mark Zuckerberg infamously dismissed as “pretty crazy” the idea that fake news on his service could have swayed the election. He later backtracked, and since then the company has reduced the amount of misinformation shared on its service, as measured by several independent studies.

Zuckerberg has also, by and large, avoided similar gaffes by conceding mistakes and delivering apologies to the public and to lawmakers.

‘Stuck with all this garbage’

But even as the company bats down one problem, others pop up. The reason for that might be baked into its DNA. And that’s not just because its business model relies on as many people as possible using it as much as possible, leaving behind personal details that can then be targeted by advertisers.

“Almost everything Facebook has designed has been designed for good people. People who are nice to each other, who have birthdays to celebrate, who have new puppies and generally like to treat others well,” said Siva Vaidhyanathan, director of the Center for Media and Citizenship at the University of Virginia. “Basically Facebook is made for a better species than ours. If it were made for golden retrievers, everything would be great.”

But if just 1% of the 2.4 billion people on Facebook want to do terrible things to others, that’s 24 million people.

“Every couple of weeks, we hear about Facebook knocking down troublesome pages, making promises about hiring more people, building AI and so on,” Vaidhyanathan said. “But at Facebook’s scale, none of that will matter. We are basically stuck with all this garbage.”

Chris Hughes, a co-founder of Facebook, called for a breakup of the social media giant in a Thursday op-ed. Vaidhyanathan also thinks strong government regulation could be the answer, such as laws that “limit companies’ ability to suck up all our data and use it to target advertising.”

“We really should be addressing the back end of Facebook,” he said. “That’s what you have to attack.”

France welcomed Facebook’s Mark Zuckerberg on Friday with a threat of sweeping new regulation.

With Facebook under fire on multiple fronts, Zuckerberg is in Paris to show that his social media giant is working hard to limit violent extremism and hate speech shared online.

But a group of French regulators and experts who spent weeks inside Facebook facilities in Paris, Dublin and Barcelona say the company isn’t working hard enough.

Just before Zuckerberg met French President Emmanuel Macron in Paris, the 10 officials released a report calling for laws allowing the government to investigate and fine social networks that don’t take responsibility for the content that makes them money.

The French government wants the legislation to serve as a model for Europe-wide management of social networks. Several countries have introduced similar legislation, some tougher than what France is proposing.

To an average user, it seems like the problem is intractable. Mass shootings are live-streamed, and online mobs are spreading rumors that lead to deadly violence. Facebook is even inadvertently creating celebratory videos using extremist content and auto-generating business pages for the likes of the Islamic State group and al Qaida.

The company says it is working on solutions, and the French regulators praised Facebook for hiring more people and using artificial intelligence to track and crack down on dangerous content.

But they said Facebook didn’t provide the French officials enough information about its algorithms to judge whether they were working, and that a “lack of transparency … justifies an intervention of public authorities.”

The regulators recommended legally requiring a “duty of care” for big social networks, meaning they should moderate hate speech published on their platforms. They insist that any law should respect freedom of expression, but did not explain how Facebook should balance those responsibilities in practice.

After meeting Macron, Zuckerberg said in a Facebook post that he welcomed governments taking a more active role in drawing up regulations for the internet. He made similar remarks earlier this year but has been vague on what kind of regulation he favors.

Facebook faces “nuanced decisions” involving content that is harmful but not illegal and the French recommendations, which set guidelines for what’s considered harmful, “would create a more consistent approach across the tech industry and ensure companies are held accountable for enforcing standards against this content,” Zuckerberg said.

The regulators acknowledged that their research didn’t address violent content shared on private chat groups or encrypted apps, or on groups like 4chan or 8chan, where criminals and extremists and those concerned about privacy increasingly turn to communicate.

Facebook said Zuckerberg is in France as part of meetings around Europe to discuss future regulation of the internet. Facebook agreed to embed the French regulators as an effort to jointly develop proposals to fight online hate content.

Zuckerberg’s visit comes notably amid concern about hate speech and disinformation around this month’s European Parliament elections.

Next week, the leaders of France and New Zealand will meet tech leaders in Paris for a summit seeking to ban acts of violent extremism and terrorism from being shown online.

Facebook has faced challenges over privacy and security lapses and accusations of endangering democracy — and it came under criticism this week from its own co-founder.

Chris Hughes said in a New York Times opinion piece Thursday that it’s time to break up Facebook. He says Zuckerberg has turned Facebook into an innovation-suffocating monopoly and lamented the company’s “slow response to Russian agents, violent rhetoric and fake news.”

Earlier this month, officials from a group of 30 countries agreed to take a more coordinated approach to secure the next generation of fast mobile communication networks, known as 5G. The United States and others worry that technology companies located in countries with governments like China’s could be subject to state influence, making the networks insecure. Elizabeth Lee reports on the security concerns over 5G, and what it means to consumers.

A senior U.S. State Department official said there is no need for President Donald Trump to sign an executive order to explicitly ban Chinese telecommunication company Huawei from taking part in the buildout of the U.S. 5G networks.

The four largest U.S. telecom carriers — Verizon, AT&T, T-Mobile and Sprint — have agreed not to use Huawei in any part of their 5G networks, said Ambassador Robert Strayer, deputy assistant secretary of state for cyber and international communications and information policy.

Strayer spoke with VOA about U.S. 5G policy and security concerns over Huawei. He also said the United States will only use trusted vendors, including South Korea’s Samsung, Sweden’s Ericsson and Finland’s Nokia, in the buildout of the U.S. 5G networks.

The following is an edited excerpt of the interview:

VOA: VOA broadcasts to many countries in Africa and Asia. These are places eager to develop their economies with high-tech communications. What does the U.S. say to those countries, which are eager for 5G and see the most attractive equipment and financing packages for those networks are all Chinese? If countries resist the Huawei offer, how many years back does that set their 5G networks? What would be the alternatives?

Deputy Assistant Secretary Robert Strayer: All around the world, we’re all very excited to see the promise of 5G technology. It’s going to empower things like telemedicine, autonomous vehicles, autonomous manufacturing, and including autonomous transportation networks in general.

So it’s going to be very important that network be incredibly secure because of all the critical infrastructure that’s going to ride on top of it. We know that there are a number of vendors besides Chinese technology vendors that are providing the equipment, the underlying infrastructure for 5G networks.

Those include Samsung in South Korea, Ericsson in Sweden and Nokia in Finland. So we believe those are trusted vendors.

We have grave concerns about the Chinese vendors because they can be compelled by the National Intelligence Law in China as well as other laws in China to take actions that would not be in the interests of the citizens of other countries around the world. Those networks could be disrupted or their data could be taken and be used for purposes that would not be consistent with fundamental human rights in those countries.

VOA: But it’s going to be a difficult choice. China is offering a great deal, in some cases 0% interest loans, 20-year payment plans, and what are the alternative plans like? Is there an analogy that you have that can show how turning down that kind of offer for something like 5G is actually in their long-term interest?

Strayer: We think that there should be commercially reasonable terms applied to financing deals. There’s obviously private financing available from telecom companies, but there are also a number of multinational, multilateral development banks providing potential sources of financing for infrastructure deals around the world.

We don’t think that countries need to adhere to, be left with only the predatory lending terms that are often offered by the Chinese Development Bank and other financing mechanisms that the Chinese companies are offering. Zero percent interest for 20 years is not commercially reasonable. It comes with huge strings attached. In fact, many of these things aren’t even transparent enough for countries to know what they’re signing up to.

We’re encouraging countries to think carefully about how they will move into 5G, make sure that they’re applying and signing up to financing terms that are commercially reasonable and ones that they can pay back in the long term.

We know of stories, of course, of ports being used as collateral in some of these financing deals, so countries could lose access to their very critical infrastructure under the terms of some of these deals. So we think that while 5G has huge promise and we should move quickly to it, we’re not in any way slowing ourselves down by going with vendors that are more trustworthy, and under financing conditions that are probably concessionary but are not at the level of some of these deals that are in no way reasonable in any type of commercial sense.

VOA: If Washington is asking other countries to ban Huawei from their 5G networks, why hasn’t the U.S. done so? I mean, the president has not signed an executive order on a comprehensive ban on Huawei, not just in the government, but in the private sector as well. Is the U.S. credibility at stake? How certain are you that the U.S. will ban Huawei equipment from its 5G network?

Strayer: So in our view, we don’t need to have a legal mechanism to ban Huawei in our private sector networks. The four largest U.S. telecom carriers have already agreed that they will not use Huawei or ZTE in any part of their 5G networks and they’re not using it in their 4G networks. So we don’t think that we need a legal tool to force them to do so. In addition, last year in the National Defense Law that was enacted at the end of the year, the government was prohibited — our U.S. government is prohibited from using these high-risk vendors.

VOA: Chinese Vice Premier Liu He is coming to Washington this week for the latest round of trade negotiation with the U.S. There are allegations against Huawei for stealing U.S. intellectual property. How should Huawei and 5G be discussed in the bilateral trade talks? Could they be hurdles for the two nations to reach a deal?

Strayer: I just want to be very clear that everything we’re talking about with countries around the world is about a national security threat that we see facing now, and that we think could have significant economic implications for them as well.

We are not talking about this in the context of trade. And I would just mention, too, that the concerns we have about Huawei that are well-documented are related to corruption, related to the theft of intellectual property, and related to defying sanctions, and using basically money-laundering schemes, have raised great concern about that company itself, but they’re not part of our trade discussions.

VOA: Is the U.S. lagging China in developing 5G infrastructure?

Strayer: No. We think we’re leading the world. By the end of this year, we’ll have 90 trials rolled out across the United States. We’ve already seen them being rolled out by Verizon and AT&T. We think we are actually leading the world in this field and we’re using only vendors from those three countries I mentioned that are trusted vendors, not the ones in China.

VOA: Thank you for talking to VOA.

Strayer: Thank you.

U.S. law enforcement officials announced on Wednesday the indictment of two Israeli operators of a website that referred hundreds of thousands of users to underground internet marketplaces to purchase drugs, weapons and other illegal products.

 

Tal Prihar, 37, an Israeli citizen living in Brazil, and Michael Phan, 34, who lives in Israel, were indicted by a federal grand jury in Western Pennsylvania with money laundering in connection with operating DeepDotWeb, a website that served as a gateway to the Darkweb, the internet’s dark underbelly where users can purchase and exchange illegal products.

Prihar was arrested by Paris authorities on Monday and faces likely extradition to the U.S. Phan was arrested Monday in Israel and faces charges there. Prosecutors declined to say whether they’ll seek Phan’s extradition to the U.S. 

 

The two Israeli nationals operated the DeepDotWeb from 2013 to late last month when it was taken down by the FBI, collecting more than $15 million in commissions for directing users to various marketplaces such as the now defunct AlphaBay. 

 

The users, in turn, purchases hundreds of millions of dollars worth of illegal drugs, firearms, malicious software, hacking tools, and stolen financial information and credit cards, according to prosecutors. 

 

About 24 percent of all orders on AlphaBay, which was one of the largest Darkweb marketplaces before it was seized by the FBI in 2017, were associated with an account created through a referral link provided by DeepDotWeb. 

 

Hyperlinks to Darknet marketplaces

Scott W. Brady, the U.S. attorney for Western Pennsylvania, said the DeepDotWeb’s takedown represents a major blow to the Darknet economy. 

 

“This is the single most significant law enforcement disruption of the Darknet to date,” Brady said at a press conference in Pittsburgh. “While there have been successful prosecutions of various Darknet marketplaces, this prosecution is the first to attack the infrastructure supporting the Darknet itself.” 

Darknet marketplaces operate on Tor, a computer network that facilitates anonymous communication and transactions over the internet. Tor marketplaces can’t be found via a Google search. To access a marketplace, a user needs the site’s exact .onion url, a top level domain suffix designating an anonymous service reachable via the Tor network.

To address this problem, DeepDotWeb provided pages of hyperlinks to various marketplaces such as AlphaBay Market and Hansa Market, allowing users to navigate the marketplaces and collecting a commission each time a user made a purchase. 

Google’s self-driving car spinoff, Waymo, is teaming up with Lyft in Arizona to attempt to lure passengers away from ride-hailing market leader Uber.

The alliance announced Tuesday will allow anyone with the Lyft app in the Phoenix area to summon one of the 10 self-driving Waymo cars that will join the ride-hailing service by end of September.

Waymo’s robotic vehicles will still have a human behind the wheel to take control in case something goes awry with the technology. But their use in Lyft’s service could make more people feel comfortable about riding in self-driving cars.

Self-driving to a profit

Both Lyft and Uber consider self-driving cars to be one of the keys to turning a profit, something neither company has done so far. Meanwhile, Waymo has been slowly expanding its own ride-hailing service in the Phoenix area that so far has been confined to passengers who previously participated in free tests of its self-driving technology.

“We’re committed to continuously improving our customer experience, and our partnership with Lyft will also give our teams the opportunity to collect valuable feedback,” Waymo CEO John Krafcik wrote in a blog post.

Lyft President John Zimmer described the Waymo partnership as “phenomenal” in a Tuesday conference call. Uber didn’t respond to a request for comment.

The new threat to Uber is emerging as the San Francisco company pursues an initial public offering of stock that could raise $9 billion when the deal is completed later this week. Lyft raked in more than $2 billion in its own IPO in March, only to see its stock fall nearly 20% below its offering price amid concerns about its ability to make money, a challenge magnified by another loss of $1.1 billion during the first three months of the year.

Waymo invests in both

Waymo’s corporate parent, Alphabet Inc., is in line to be among the biggest winners in Uber’s IPO just as it was in the Lyft IPO. Alphabet owns a 5% stake in Uber that will be worth as much as $3.6 billion if Uber realizes its goal of selling its stock for as much as $50 per share. It also holds a 5% stake in Lyft that is currently worth $761 million.

Despite their financial ties, Waymo and Uber have had an acrimonious relationship since becoming entangled in a thorny case of alleged high-tech theft.

Waymo accused Uber of orchestrating a scheme to steal some of its autonomous driving technology. That came after Uber’s former CEO Travis Kalanick began to suspect Waymo was planning to use its self-driving cars in a rival ride-hailing service.

The two sides settled that dispute last year in a deal that required Uber to give Alphabet another bundle of stock that was worth $245 million at the time the truce was reached.

The agreement also requires Uber to submit to reviews by a software expert to ensure it isn’t misusing any of Waymo’s technology in its effort to build its own self-driving cars, a process that recently uncovered some potentially “problematic” issues, according to discloses made as part of Uber’s IPO. Uber warned the problems could require it to pay a licensing fee to Waymo or delay its efforts to introduce self-driving cars in its service.

Google CEO Sundar Pichai is expected to showcase much-anticipated updates to the company’s hardware lines and artificial intelligence.

Google will also likely address privacy updates as concerns about data sharing continue to plague the tech industry. Facebook dedicated much of its own conference last week to addressing privacy.

Rumors suggest that Google may unveil a mid-range Pixel phone as a cheaper option to the flagship model currently on sale for $800.

Pichai has a keynote scheduled Tuesday at the company’s annual I/O conference for software developers in Mountain View, California.

Google says more than 7,000 developers will attend. The conference is focused on updates for the computer engineers that build apps and services on top of Google technology. I/O has also become a stage to announce new consumer products.

Alphabet’s Google is set to roll out a dashboard-like function in its Chrome browser to offer users more control in fending off tracking cookies, the Wall Street Journal reported on Monday, citing people familiar with the matter.

Cookies are small text files that follow internet users and are used by advertisers to target consumers on the specific interests they have displayed while browsing.

While Google’s new tools are not expected to significantly curtail its ability to collect data, it would help the company press its sizable advantage over online-advertising rivals, the newspaper said.

Google’s 3 billion users help make it the world’s largest seller of internet ads, capturing nearly a third of all revenue, ahead of rival Facebook’s 20%, according to research firm eMarketer.

Total digital ad spending in the United States will grow 19%  to nearly $130 billion in 2019, according to eMarketer.

Google has been working on the cookies plan for at least six years, in stops and starts, but accelerated the work after news broke last year that personal data of Facebook users was improperly shared with Cambridge Analytica.

The company is mostly targeting cookies installed by profit-seeking third parties, separate from the owner of the website a user is actively visiting, the Journal said.

Apple in 2017 stopped majority of tracking cookies on its Safari browser by default and Mozilla’s Firefox did the same a year later.

Google did not immediately respond to a Reuters’ request for comment.

When Google launched its now distinctive digital assistant in 2016, it was already in danger of being an also-ran.

At the time, Amazon had been selling its Echo smart speaker, powered by its Alexa voice assistant, for more than a year. Apple’s Siri was already five years old and familiar to most iPhone users. Google’s main entry in the field up to that point was Google Now, a phone-bound app that took voice commands but didn’t answer back.

Now the Google Assistant – known primarily as the voice of the Google Home smart speaker – is increasingly central to Google’s new products. And even though it remains commercially overshadowed by Alexa, it keeps pushing the boundaries of what artificial intelligence can accomplish in everyday settings.

For instance, Google last year announced an Assistant service called Duplex, which it said can actually call up restaurants and make reservations for you. Duplex isn’t yet widely available yet outside of Google’s own Pixel phones in the U.S. Alexa and Siri so far offer nothing similar.

Google is expected to announce updates and expansions to its AI Assistant at its annual developer conference Tuesday.

Although voice assistants have spread across smartphones and into cars and offices, they’re currently most commonly found in the home, where people tend to use them with smart speakers for simple activities such as playing music, setting timers and checking the weather. Amazon’s Echo devices maintain a strong lead in the market, according to eMarketer ; the firm estimates that 63% of all U.S. smart speaker users will talk to an Amazon device this year, compared to 31% that will use Google. Apple’s HomePod is a mere afterthought, lumped in the “other” category which has a combined 12%.

More broadly, though, the competition is much more difficult to assess. Google claims the Assistant is now available across more than a billion devices, although many of those are smartphones whose owners may never have uttered the Assistant’s wake-up phrase, “OK Google.”

Google Assistant doesn’t record users commands by default – differing from Alexa – but recording must be turned on to access some of Assistant’s features, including a popular one that allows it to recognize different users by voice.

​Amazon and Google may one-up each other on different metrics, but the real measurement is how well they’ve achieved those own goal, said Gartner analyst Werner Goertz.

Amazon’s deep ties in shopping make Alexa the go-to assistant for adding items to your grocery list or putting in a quick re-order of dish soap. Google’s decades of deep search technology make it the leader in looking up or answering questions you might have and personalizing its responses based on what else Google knows about you from your previous searches, your movements or your web browsing.

All that, of course, reinforces Google’s key advertising business, which is based on showing you ads targeted to your interests.

At first, the Assistant on Home mostly just acted as a vocal search engine; it could also carry out a few additional tasks like starting your Spotify playlists. Over time, however, it has added dozens of languages, partnered with more than 1,500 smart home companies to control lights, locks and TVs and learned to identify members of any given household by voice.

It’s also expanded the number of apps and other companies it works with and moved into Google Maps as a way to send text messages while driving.

Both Google and Amazon plan further expansions. Last year, Amazon unveiled a number of home gadgets with Alexa built in, including a “smart” microwave. At the CES gadget show this year, it showed off a phone-connected device that brings Alexa to cars. 

Google countered with updates to its expanding Android Auto system, which got Assistant capability last year.

As Assistant and Alexa get smarter, faster and more personalized, analysts expect their reach to become broader and more ubiquitous. The speakers, said eMarketer analyst Victoria Petrock, are “getting people used to talking to their devices.” Eventually, she says, if you can speak to your microwave and TV and lights directly, you won’t need the speakers – except maybe to play music.

In these emerging areas Google is hoping to outflank rivals with its strong inroads with Android smartphones and cars. But it faces competition in many of these areas not just from Amazon, but also Apple and Microsoft.

Google I/O kicks off at 10 a.m. Tuesday in Mountain View, California. The company is expected to announce a less expensive Pixel phone and updates to its smart home devices.

Microsoft is announcing an ambitious effort intended to make U.S. voting more secure and verifiable.

The company will offer free open-source software that several top U.S. elections vendors say they will explore incorporating into their voting equipment.

 

The software kit is being developed with Oregon-based Galois, which is separately creating a secure voting system prototype under contract with the Pentagon’s advanced research agency, DARPA.

 

The Microsoft kit is dubbed “ElectionGuard.” It was announced Monday by CEO Satya Nadella at a Seattle developer’s conference.

 

Microsoft says the kit it will be available beginning this summer with prototypes ready to pilot for next year’s general elections.

 

 

Virtual Reality had a fantastic year in 2016, with the release of several anticipated VR glasses, including the Oculus Rift and the HTC Vive. Gaming and technology fairs presented the new toys proudly, but the boom quickly declined, leaving the technology to only niche applications. Now, a southeastern Chinese city has opened an entertainment park that intends to show VR’s potential as a future technology. Markus Meyer-Gehlen reports.

An innovative Czech company is changing the way we think of light, by capturing and transforming it in ways that save money and generally make the world a more beautiful place. VOA’s Kevin Enochs reports.

Lightweight, cheap to make and easier to send into outer space. Academics in the U.S. are developing an inflatable robot with money from the American space agency, NASA. NASA says the blow-up technology can handle the cosmos and existence back here on Earth. Arash Arabasadi has more.