Amazon’s Echo Speakers Are Spontaneously Laughing—And Users Are Spooked

Amazon is trying to stop its Amazon Echos, powered by the Alexa voice-activated assistant, from suddenly laughing.

The online retail giant told to tech news publication The Verge on Wednesday that it is aware that some Echo Internet-connected speakers have a laughing problem and is “working to fix it.”

Amazon’s (amzn) acknowledgement of the issue comes after several people have reported on Twitter and Reddit that their Amazon Echo speakers have started laughing, for no apparent reason. People typically activate their Echo speakers by saying the word “Alexa,” which triggers the device to listen and respond to commands like changing the volume.

The latest laughing is causing some customers to feel unsettled and confused. Other say it’s spooky, like something out of a horror film.

Although third-party devices like some HP Inc. personal computers and certain modems work with Alexa, the problem appears to be limited to the Amazon Echo and the smaller Echo Dot.

A Reddit user also described problems with the Amazon Echo Dot two weeks ago:

We just added the echo dots two months ago. The dot we have in the master bath has twice now randomly played a track of a woman laughing at about 10 p.m. the first time I thought the fire tv was sending audio through it since I had been trying to sync them up to the tv, but tonight was completely random. No indication on the app that the device heard any command. We had the dot laugh several times and it wasn’t the laugh Alexa produces, but definitely sounded like a canned laugh, not like someone laughing live.

It’s unclear what’s causing the errors and Amazon’s brief statement did not say when the problem would be fixed or if other Alexa-enabled devices are also affected.

Get Data Sheet, Fortune’s technology newsletter.

Fortune contacted Amazon for more details and will update this story if it responds.

Google Just Indexed Millions of ‘Life Magazine’ Photos Using Artificial Intelligence

Google has used its artificial intelligence to automatically index millions of photographs from the defunct Life Magazine.

The search giant, which debuted a website for the photo project on Wednesday, said it was able to categorize over 4 million iconic Life Magazine photographs without human help. After clicking on a particular label like “skateboarding,” for example, users are shown photos of people performing skateboard tricks along with Wikipedia’s definition of the sport.

Google uses deep learning technology to help its computers better understand objects like dogs or cats in pictures and make its image search tool more efficient to people wanting to see a particular image.

Compared to the core Google search, the photo project’s website is slow to load, especially when there are thousands of images assigned to a particular label like “road.”

Although Google has hosted a Life Magazine archive since 2008, the new website makes using it easier. Searching Life Magazine’s photojournalism for ballet brings up relevant photos alongside the name of the photographer who took the picture, the title of the photo like “Nutcracker Ballet,” and what in the photo Google’s computers were able to recognize, like a dancer.

Although usually accurate, Google’s technology does have some hiccups that highlight some of AI’s current limitations. For instance, under the “skateboarding” label is a photo that is shown sideways of a man wearing a top hat and holding a cane who is performing what looks like a campy musical number. For unknown reasons, the computers mistakenly thought it saw a skateboard in the image.

Additionally, the computer that created the index has come up with some odd categories that human editors likely wouldn’t have. One such label is “identity document,” which brings up photos of people’s passports and railroad tickets as well as a photo of musician Paul McCartney holding a Grammy plaque. It’s likely the computers saw similarities between a typical document and McCartney’s Grammy plaque, which in this case resembled more of a small, commemorative plate than the conventional 3-dimensional Grammy statue.

Get Data Sheet, Fortune’s technology newsletter.

Another label, “concrete,” highlights thousands of photos, some of which appear random. There’s a picture of an old tombstone, a photo of a marine boot camp, a picture of a person hunting a snake in a small enclosure, and a photo of the “Roosevelt Raceway” that shows a man—incorrectly identified by the computer as a fisherman— sweeping. While there indeed seems to be some sort of concrete object involved with each photo, it’s still an unconventional way to group all the pictures together considering they are displaying very different subjects.

In addition to the misidentified photos, the technology also failed to create labels for some obvious topics. They include some of Life Magazine’s most iconic photos.

For instance, the computers did not create a label for the “Vietnam War,” a subject that should include some of Life Magazine’s best known and most powerful photos. The computer did create is a “war” label, but it appears to have grouped over 23,000 photos, making searching for a particular picture difficult.

Cryptocurrencies are risky for consumers, says BoE's Haldane

LONDON (Reuters) – Cryptocurrencies pose a risk to British consumers, though not to the financial system as a whole, the Bank of England’s chief economist, Andy Haldane, said on Tuesday.

FILE PHOTO: A collection of Bitcoin (virtual currency) tokens are displayed in this picture illustration taken December 8, 2017. REUTERS/Benoit Tessier/Illustration/File Photo

“There’s lots of potential risks there, one of which is the danger to the consumer from buying into this stuff,” Haldane said in a BBC television interview.

Bitcoin BTC=, the best known cryptocurrency, soared in value from around $1,000 at the start of 2017 to almost $20,000 in mid-December, before tumbling below $6,000 last month and then staging a partial recovery.

Haldane’s concerns are similar to those expressed by BoE Governor Mark Carney in a speech on Friday, and previously by Britain’s Financial Conduct Authority.

Many global regulators have warned about cryptocurrencies this year and China has banned financial institutions from processing them. Carney said this would be a step too far, given the potential of the underlying technology to improve payments and asset clearing and settlement.

Haldane said the BoE continued to monitor cryptocurrencies, and that at less than 1 percent of total global wealth, they did not pose a big danger to the world’s financial system.

But asked if he would invest in cryptocurrencies himself, Haldane said he was very risk averse, and would not.

Reporting by David Milliken; Editing by James Dalgleish

Pennsylvania Sues Uber Over Data Breach Disclosure

Uber faces more potential legal consequences for waiting to make public a major hack until a over a year after it happened. The Pennsylvania Attorney General filed a lawsuit against Uber Monday for violating the state’s data breach notification law, which says hacks should be disclosed within a “reasonable” time frame. Uber didn’t merely keep quiet about the massive breach; it reportedly paid a $100,000 ransom to the perpetrators in exchange for their silence. And while experts say Uber will likely settle the case, it may be just the latest in a cascade of similar lawsuits.

The stolen Uber data included the names and driver’s license information of around 600,000 drivers—including at least 13,500 from Pennsylvania—as well as data belonging to 25 million users in the US. It impacted over 57 million people in total. “Uber violated Pennsylvania law by failing to put our residents on timely notice of this massive data breach,” Josh Shapiro, the states’s attorney general, said in a statement. “Instead of notifying impacted consumers of the breach within a reasonable amount of time, Uber hid the incident for over a year and actually paid the hackers to delete the data and stay quiet.” Under Pennsylvania’s data breach notice law, the attorney general may seek fines up to $1,000 for each violation, leading to a maximum penalty of $13.5 million for Uber.

Pennsylvania’s joins a growing line of lawsuits against the ride-share company. Both Washington state, and cities including Los Angeles and Chicago filed suits when the breach was first made public by the company’s new CEO Dara Khosrowshahi in November. Two class-action lawsuits were also filed in California days after the breach was first disclosed. Attorneys general from New York, Missouri, and Connecticut have also said they would look into the breach. Forty-eight states, (excluding South Dakota and Alabama) currently have laws on the books regulating how and when a data breach must be disclosed.

“Since starting on this job three months ago, I’ve spoken with various state and federal regulators in connection with the data breach pledging Uber’s cooperation, and I personally reached out to Attorney General Shapiro and his team in the same spirit a few weeks ago. While I was surprised by Pennsylvania’s complaint this morning, I look forward to continuing the dialogue we’ve started as Uber seeks to resolve this matter,” Tony West, Uber’s chief legal office said in a statement. “We make no excuses for the previous failure to disclose the data breach. While we do not in any way minimize what occurred, it’s crucial to note that the information compromised did not include any sensitive consumer information such as credit card numbers or social security numbers, which present a higher risk of harm than driver’s license numbers. I’ve been up front about the fact that Uber expects to be held accountable; our only ask is that Uber be treated fairly and that any penalty reasonably fit the facts.”

The Pennsylvania lawsuit is also the first to cite a Senate hearing in February where John Flynn, Uber’s chief information security officer, testified in front of the Committee on Commerce, Science, and Transportation about the hack. Uber initially said the payment it made to the hackers responsible for the breach was not a ransom, but simply a payout under its existing bug bounty program, a system many tech companies deploy to reward security researchers for bringing vulnerabilities to their attention. But during the hearing, Flynn acknowledged that the agreement made with the perpetrators—as well as the $100,000 payment—were not typical for its bug bounty program, which usually compensates researchers only a couple thousand dollars.

“The fact that this was a multistep malicious intrusion, a downloading of data, and extortionate demands means this wasn’t consistent with the way that [the bug bounty] program normally operates,” Flynn testified. He also said that Uber “made a misstep not reporting to law enforcement.”

William McGeveran, a professor at University of Minnesota Law School who specializes in data privacy law, said it’s possible Uber will settle with Pennsylvania for a fraction of the total $13.5 million fine, and take on commitments to ensure a similar breach doesn’t happen in the future. “In these settlements many times regulators care more about fixing the problem than about being punitive,” says Mcgeveran. But more suits could follow from other states, especially because Flynn’s statements before the Senate committee provide state prosecutors with more evidence to work with.

“Given the alleged facts in this case, it wouldn’t surprise me at all to see more lawsuits,” says Woodrow Hartzog, a law and computer science professor at Northeastern University who studies privacy and data protection issues. “Oftentimes you will have state attorneys general that might even work together if that appears to be the best course of action. They’ll probably be using the facts in this case as an example of how not to respond to a data breach.”

Uber has also already faced disciplinary action from federal regulators twice, once for a separate hack in 2014 that exposed the information of 100,000 drivers, and once for misleading drivers about how much money they could make. The FTC said in November that it was also evaluating the “serious issues” raised by the 2016 breach.

Uber has yet to pay any fines to the federal government, and won’t have to if it makes good on its promises to protect its drivers’ and customers’ privacy. The agreement between the FTC and Uber lasts 20 years. If the FTC decides that the 2016 breach is considered a violation of that agreement, the ride-hailing company could face expensive consequences. In 2012 for example, the FTC fined Google $22.5 million for violating its 2011 settlement.

For now, no federal law exists requiring companies disclose a data breach within a certain time frame. But since nearly every state has a data breach law, Uber could still face a patchwork of further lawsuits. Some lawmakers are also pushing for federal legislation. In December, Democratic senator Bill Nelson introduced the Data Security and Breach Notification Act, which would require companies to report breaches within a month, or face up to five years in prison.

Federal laws punishing companies for failing to notify about a breach wouldn’t necessarily improve protections for consumers, however. “I would be skeptical of the claims that a unified data security protection law are going to provide clarity and better data protection at the same time,” says Hartzog, who has testified before Congress about data breach legislation. “A movement to have a single unified standard among the United States would be seen as an opportunity to water down those requirements.”

State laws also give attorneys general the chance to act if they perceive the Federal Trade Commission to be not aggressive enough. “I think we’re going to see more activity by state attorneys general in privacy and security cases because it’s not clear how much the FTC is going to do under its current management compared to previous,” says McGeveran. “These states have a better argument because they have specific requirements that you notify about a breach.”

Besides, it’s not hard for a major corporation like Uber to juggle multiple state regulations at once, especially because the ones governing breach disclosure mandate the same things. “Many of them are quite similar in their requirements, many of them have the same deference to industry standards,” says Hartzog. It’s far harder to navigate, say, every state’s regulations on taxis.

Uber Issues

Reddit CEO Steve Huffman Acknowledges Users Shared Russia Propaganda

Russian trolls used Reddit to spread propaganda in prelude to the 2016 U.S. presidential election, Reddit said on Monday.

Reddit CEO Steve Huffman said in a post on Reddit that the company removed a “few hundred accounts” from its social media service that it believed were linked to Russian-based entities that had spread misleading information.

Huffman did not identify the groups, but the Daily Beast reported last week that members of the Russia-based Internet Research Agency had spread misinformation on Reddit’s various message boards as well as on Tumblr blogging service. The IRA was one of three Russian groups identified in a recent Justice Department indictment alleging that Russian individuals had spread fake news and propaganda through popular social networking and messaging services like Facebook (fb) and Twitter (twtr) in an effort to exacerbate existing divisions in the U.S..

“As for direct propaganda, that is, content from accounts we suspect are of Russian origin or content linking directly to known propaganda domains, we are doing our best to identify and remove it,” Huffman wrote. “The vast majority of suspicious accounts we have found in the past months were banned back in 2015–2016 through our enhanced efforts to prevent abuse of the site generally.”

Huffman said that there’s not much evidence that Russian trolls bought online ads on Reddit to disseminate propaganda, as they are alleged to have done on Facebook and Google.

“We don’t see a lot of ads from Russia, either before or after the 2016 election, and what we do see are mostly ads promoting spam and ICOs,” Huffman said.

Huffman also said that thousands of U.S.-based Reddit users may have unwittingly promoted Russian propaganda on the service. He cited the fact that these Reddit users endorsed the postings of @TEN_GOP Twitter account, which they thought were linked to a real Republican group but was actually a “Russian agent.”

Get Data Sheet, Fortune’s technology newsletter.

“I wish there was a solution as simple as banning all propaganda, but it’s not that easy. Between truth and fiction are a thousand shades of grey,” Huffman wrote. “It’s up to all of us—Redditors, citizens, journalists—to work through these issues.”

The Senate Intelligence Committee now wants more information from Reddit about the possibility that Russia may have exploited the service, the Washington Post reported Monday. The committee also plans to hold a briefing with Tumblr, the newspaper said citing an unnamed source.

And The Oscar Goes To… 

On days like this predictions abound. And most often those predictions come from the pundits, the ones “in the know.” But what if the real power of prediction is not in the opinions and the bias of pundits but in the data?

One the most amazing things about the internet is its ability to quickly determine the sentiment of large populations just by listening to what people are talking about. Of course, listening is nothing new. Anyone building a successful business will tell you that the single most important skill to develop and info into a culture is that of listening to the customer. 

What if I told you that was wrong? Anathema, right? Customers always come first! That’s only partially true. The problem with always putting the customer first is that you build an echo chamber which reports that same old biased view of what you should be doing to grow your business based on what you have been doing, while growth comes not only from customer but from those who are not yet customers. 

Companies often try desperately to hang onto customers when their overall business is shrinking and suffering, They put in place loyalty programs and incentives for customers to stay, but they are oblivious to the sentiment of the larger market which is fleeing to other alternatives. It’s like trying to give cabin upgrades to passengers of the Titanic–while it’s sinking!  

Instead, what if you could listen to the entire market, current and potential, in order to direct your resources and align your decisions with where the growth was? What if you could predict the many ways in which the marketplace is changing but which no individual customer, focus group, marketing genius, or existing community of customers could adequately express?

That’s precisely the objective of social listening; to understand the collective sentiment of a marketplace based on what the data is telling your. It sounds simple but it’s amazing how few companies are doing it. Why not? Because we all want to believe that we are smarter than the market; that the data has nothing on us. Because, if it did, what would our value be?

Which brings me to the Oscars. (You knew I’d get there eventually!)

Using a listening analytics platform, Sprout Social captured data around the three major Oscar categories of best picture, best actor in a leading role, and best actress in a leading role in order to project tonight’s Oscar winners. Although this is just for fun, it’s a great illustration of how powerful social sentiment can be in understanding a market’s perspective.

In each category they scoured the web for the number of mentions and the positive versus negative mentions. The results are in some cases straight forward and in others fascinatingly close. 

For example, the data shows that Call Me By Your Name is the projected winner among fans, garnering 152,880 total mentions, 64,758 positive mentions, 18,095 negative mentions and 46,663 net positive mentions (that’s the number of positive mentions less the negative mentions.)  The Shape of Water and Lady Bird follow close behind with 48,039 and 34,268 positive mentions respectively. However, Dunkirk had more total mentions than Lady Bird, but just about 7,000 fewer net positive mentions. Although Get Out had more net positive mentions than Dunkirk! 

None of that is likely to usurp Call Me by Your name, but from a marketing standpoint it provides insight into how close sentiment is about movies that may not be getting an Oscar but are none-the-less neck and neck in terms of popularity. 

Even more fascinating is the ridiculously close net positive ratings for the category of Best Actress in a Leading Role as compared to the category of Best Actor in a Leading Role. Make your own inferences here but the bifurcation of opinions is at the very least a fascinating look at how polarized sentiment can become. 

Best Motion Picture TOTAL MENTIONS POSITIVE MENTIONS NEGATIVE MENTIONS NET POSITIVE MENTIONS
Call Me By Your Name  152,880 64,758 18,095 46,663
The Shape of Water 115,578 48,039 12,304 35,735
Lady Bird  64,063 34,268 7,249 27,019
Dunkirk  73,586 33,085 12,988 20,097
Get Out  56,196 32,136 9,802 22,334
Best Actor in a Leading Role
Daniel Kaluuya (Get Out) 89,552 37,152 4,698 32,454
Timothée Chalamet (Call Me by Your Name) 28,527 16,405 1,785 14,620
Gary Oldman (Darkest Hour) 15,057 8,337 2,573 5,764
Best Actress in a Leading Role
Margot Robbie (I, Tonya) 2,007 593 589 4
Meryl Streep (The Post) 549 196 183 13
Saoirse Ronan (Lady Bird) 404 188 34 154
Frances McDormand (Three Billboards) 202 81 22 59
Sally Hawkins (The Shape of Water) 131 70 15 55

*Data provide by Sprout Social

So, will the data foretell tonight’s winners? The practical side of this specific scenario is that the 6,000 members of the Academy of Motion Pictures and Sciences who vote on the Oscars can do whatever they want, regardless of what the data says. Are they likely to reflect the broader demographic captured in the table above? Probably, but not necessarily. 

Whatever the case, we are at a point when we need to spend more time listening to the data and less time in our echo chambers.

SpaceX is Days Away From a Major Milestone for the Falcon 9 Rocket

This Tuesday’s planned launch of the Hispasat 30W-6 from Kennedy Space Center will be a landmark for Elon Musk’s SpaceX. If everything goes according to plan, it will be the 50th time the company’s Falcon 9 rocket successfully heads into the stratosphere since its inaugural flight in June 2010.

According to Ars Technica’s Eric Berger, who highlighted the imminent landmark, the Falcon 9 should make it to 50 launches faster than some comparable programs. The United Launch Alliance Atlas V rocket took nine years and seven months to hit that mark, while the space shuttle program launched 50 times in its first 11 years and 5 months.

Get Data Sheet, Fortune’s technology newsletter.

Of course, not every Falcon 9 mission has been a success. The rocket’s 19th attempted launch, a space station resupply mission, failed in June 2015. And a particularly embarrassing launchpad explosion destroyed an expensive satellite, partly underwritten by Facebook, in the fall of 2016.

But 2017 was a banner year for the company, with a record 18 Falcon 9 launches, and no major launch mishaps. This year is also off to a good start, with three successful Falcon 9 launches. One of those missions ended with the reported loss of the secretive Zuma satellite, but SpaceX has maintained that wasn’t its fault. On top of that, the long-awaited Falcon Heavy launched successfully in early February.

Assuming Tuesday’s launch is similarly uneventful, SpaceX’s 2018 will be jam-packed. Around a dozen Falcon 9 launches and two commercial Falcon Heavy launches are already scheduled, but there are other payloads waiting, yet still unscheduled. SpaceX is aiming to open a new launch facility in south Texas before the end of the year, so things could get downright hectic.

Drunk New Jersey Man Accidentally Takes 300-Mile, $1,600 Uber Ride

New Jersey native Kenny Bachman got more than he bargained for when he climbed into an Uber last Friday, after a night of partying with friends in Morgantown, West Virginia. Bachman was staying on the campus of West Virginia University, and thought he had summoned a car to take him there.

But Bachman apparently blacked out, and when he came to, he was instead halfway back to New Jersey.

“I just woke up,” Bachman told NJ.com, “And I’m thinking, ‘Why the f—- am I in the car next to some random ass dude I don’t even know?”

Bachman had little choice but to complete the 300-mile ride. But when they arrived at his home in Gloucester County, he saw the total: $1,635.93.

Get Data Sheet, Fortune’s technology newsletter.

Any 300-mile Uber ride would probably be pricey, but Bachman’s was especially steep because he accidentally ordered an UberXL, and because surge pricing was in effect, nearly doubling the fee.

Bachman initially contested the fare, claiming to NJ.com that he would never have requested a ride all the way home, and alleging that the Uber driver must have tampered with his phone.

It’s not hard to imagine, though, how an impaired customer might mistakenly summon an Uber to take them “home” to New Jersey using the Saved Places feature, without having to manually input the (wrong) address. That may be what Uber, which confirmed the ride occurred to NJ.com, explained to Bachman – after he spoke with the company, he reportedly decided to accept the fare.

Equifax breach could be most costly in corporate history

NEW YORK/TORONTO (Reuters) – Equifax Inc (EFX.N) said it expects costs related to its massive 2017 data breach to surge by $275 million this year, suggesting the incident at the credit reporting bureau could turn out to be the most costly hack in corporate history.

FILE PHOTO: Credit reporting company Equifax Inc. corporate offices are pictured in Atlanta, Georgia, U.S., September 8, 2017. REUTERS/Tami Chappell/File Photo

The projection, which was disclosed on a Friday morning earnings conference call, is on top of $164 million in pretax costs posted in the second half of 2017. That brings expected breach-related costs through the end of this year to $439 million, some $125 million of which Equifax said will be covered by insurance.

“It looks like this will be the most expensive data breach in history,” said Larry Ponemon, chairman of Ponemon Institute, a research group that tracks costs of cyber attacks.

Total costs of the breach, which compromised sensitive data of some 247 million consumers, could be“well over $600 million,” after including costs to resolve government investigations into the incident and civil lawsuits against the firm, he said.

Equifax on Thursday reported fourth-quarter profit that topped Wall Street forecasts and disclosed that it uncovered an additional 2.4 million people whose data was stolen in the attack.

Its shares rose nearly 4 percent to $115.82 on Friday on the higher-than-expected earnings. They have lost about a quarter of their value since Equifax disclosed the incident in early September.

Equifax said in September that hackers had stolen personally identifiable information of U.S., UK and Canadian consumers, including names, Social Security numbers, birth dates, addresses driver’s license and credit card numbers.

That disclosure prompted outrage from politicians and consumer advocates around the world, a string of government probes into company and the departure of top executives.

Equifax warned in regulatory filing on Thursday that further analysis could identify more consumers or additional types of data stolen in the hack.

This year’s costs include technology and security upgrades, legal fees and free identity theft services to consumers whose data was stolen, the company said in a conference call.

Reporting by John McCrank in New York and Jim Finkle in Toronto; Editing by Chizu Nomiyama and Meredith Mazzilli

U.S. Congress to vote on allowing spectrum auction for 5G networks

WASHINGTON (Reuters) – Republican and Democratic lawmakers have reached agreement to allow for the sale of spectrum to speed up the introduction of next-generation 5G wireless networks, congressional leaders said on Friday.

FILE PHOTO: Chairman Ajit Pai speaks ahead of the vote on the repeal of so called net neutrality rules at the Federal Communications Commission in Washington, U.S., December 14, 2017. REUTERS/Aaron P. Bernstein

The U.S. House of Representatives will vote on the measure on Tuesday, leaders in both parties said.

Federal Communications Commission chairman Ajit Pai on Monday said the regulator plans new auctions of high-band spectrum starting later this year for 5G networks to improve internet services across the United States.

Wireless carriers have spent billions of dollars acquiring spectrum and beginning to develop and test 5G networks, which are expected to be at least 100 times faster than 4G networks and cut latency to less than one thousandth of a second from one hundredth of a second in 4G, the FCC has said.

Pai said he plans to hold an auction of spectrum in the 28 GHz band in November, followed immediately by an auction of spectrum in the 24 GHz band, but must first get congressional approval by May 13 to proceed.

The bill unveiled on Friday includes a technical fix to allow the auction to proceed to allow the FCC to deposit upfront payments from spectrum bidders with the U.S. Treasury. Under current law, the FCC must deposit such payments with private banks in interest-bearing accounts, but regulations make that impossible.

The bill“puts consumers first and solidifies the nation’s critical telecommunications infrastructure, giving the U.S. a global edge in the race to 5G and improving internet services,” Senators John Thune, a Republican, Bill Nelson, a Democrat, and Representatives Greg Walden, a Republican, and Democrat Frank Pallone said in a statement.

The bill includes new provisions to identify more spectrum for private sector use and reduces bureaucratic hurdles connected with building wireless networks, lawmakers said.

Verizon Communications Inc has announced it will begin its first 5G commercial rollout in Sacramento, California, this year. AT&T Inc said its first 5G commercial launches in Atlanta, Dallas and Waco, Texas, later this year.

The bill authorizes funds to address a shortfall to relocate TV broadcasters who are relocating on the spectrum after the 2017 low-band spectrum auction. The FCC and law enforcement will be better able to protect consumers from fraudulent telephone calls, according to the bill.

The biggest issue it sidesteps is so-called net neutrality. Republicans have urged Democrats to negotiate to enshrine some internet protections in the law, but set limits on the FCC’s ability to regulate internet providers. Democrats argue that net neutrality rules inacted in 2015 should remain in place.

Reporting by David Shepardson; editing by Grant McCool