Just no. Almost nobody else in the western world has this problem. How? Regulation. In Germany, if you get reported for unsolicited calling, the Bundesnetzagentur will come to you and give you a very hefty fine, per violation. Problem solved. I remember exactly four unsolicited calls in my life, two of them on a number which was published in a phone book for 15 years.
Spam calling is not something were there is net value created for society so we need an elaborated discussion of what is ok and what is not (like e.g. paid universities). It just sucks. Make it illegal, and then actually enforce that. The idea of creating a million dollar industry to create a telemarket arms race is...not good.
I suspect the US is much more of a target because there's many, many more overseas criminals that speak English than German.
[0]https://www.justfacts.com/news_poorest_americans_richer_than...
The problem is that the callers disguise their identity and in most cases are overseas beyond the reach of U.S. law enforcement.
The problem is that the US telco providers let them disguise their identity.
In Germany, there are anonymous calls, and calls where you see the caller's numbers. I've never once been called by a spoofed number.
So, it's possible to prevent that, given the political will, and the necessary investments in infrastructure.
It's a problem similar to why there are more viruses for Windows, well, it's why there's more Windows machines in the world.
Is it a real thing?
In the Netherlands, there is the 'bel me niet' (don't call me) register. If you have your number put on the list, marketeers are not allowed to call you. If they do, they risk a hefty fine.
US has dished out far more severe punishments for this stuff than the Bundesnetzagentur could https://www.justice.gov/opa/pr/24-defendants-sentenced-multi...
This is not about regulations, lots of these robocallers are straight up criminal operations who will not be swayed by fines.
If Germany has implemented some technical measures against robocalling I'd absolutely love to hear about them! I seriously doubt they have though.
In case of calls originating from another country will laws and regulation actually be able to do anything here?
[1] https://gimletmedia.com/shows/reply-all/6nh3wk/102-long-dist...
So I don't think it must have been a language barrier or a market size thing.
Effectively, a call to a mobile phone costs at least about 0.01 EUR per minute. This puts a lower bound on the cost per robocall.
I wonder how much of the difference in the number of robocalls simply comes down to economics rather than to efficiency of regulators and legal frameworks.
Also, they are often manned by call centers in places like India where English speakers are easy to find, German less so.
No-one here has identified any specific effective methods through which Germany or any other European country combats these calls, it's because they can't. Fines only keep local businesses from doing this, but the people making these calls to the US are criminal gangs based out of India.
Fines do nothing to discourage this behaviour by criminals, the fact that Germans speak German has everything to do with the lack of English robocalls.
I get no spam calls on my UK number, which is in English.
How does that work if someone spoofs the caller ID?
I'm in the UK and I get multiple spam/junk/scam calls a day. What I've noticed is if I try to call back, I'll just get an error tone because the number doesn't exist.
I really don't get the negativity, I'd love to have some system like Verified by Twilio!
Stanford is a "paid university". Harvard is too. Are you saying these institutions are a net negative to society? With free market capitalism you get the good and the bad, just like every other system in the world. The difference is you're given a choice.
1. I have a "clean" phone number I only share with friends/family. Certain prefixes and phone number actually receive next to zero spam calls. This number has received 2 calls in the years I've had it.
2. I have a phone number I share with companies / loyalty programs / etc.
3. I have a legacy phone number I've had for 10 years that receives multiple spam calls daily. For this number I've set up an incoming handler that uses Google Cloud to ask the caller a simple question (What is five plus two?), parses with the Speech Recognition API, and forwards calls with the correct answer. The rest go to voicemail. It's better to use the speech recognition API than it is to ask the caller to press a DTMF tone, in case they are in a car or are calling from a device without a telephone keypad. If I every need to, I'll add this protection to use numbers 1) and 2).
4. I have a number that rings straight to voicemail.
All numbers use the Whitepages and Nomorobo APIs from Twilio. These score the "spamminess" of calls. Calls with sufficiently high values are forwarded to voicemail.
For incoming calls, I set the caller ID for the calls to my SIP clients to indicate which number is being called, and include the spam score.
The way their process worked was:
* Big company wanted a survey completed. * Big company would want it targeted to a specific state/regional area, so would provide a list of post/zip codes.
* Survey company would translate the list of codes into phone prefixes for each area.
* Computer would randomly generate phone numbers in the area - starting from xxxx 0000 through to xxxx 9999 for operators to call.
That meant you'd get disconnected numbers, businesses, private/silent numbers, faxes, modems, etc.
So having a "clean" number didn't protect you. While a few projects were targeted to specific people/numbers - most were just random.
Were you the one making calls? How did go through the randomly generated phone numbers?
Having kids, buying a house/land, starting a business, contracting with services - every once in awhile I’ll get a high priority call that I really don’t want to be delayed (emergency at school or daycare, last minute problem with land/house offer, etc.). One time even my doggie daycare called because my dog had collapsed and they were doing CPR in the background and needed my consent to emergently administer epinephrine.
I also live in a rural area and I wonder if I tried to get a local business or contractor to return a call if they’d assume they had a wrong number if they called and a robot asked them to do a math problem. I know I’d not infrequently get confused voicemails with my Google Voice number (“Not sure I’ve got the right number but ...”).
So in the interest of not wanting to risk those rare important calls being delayed or blackholed I’m one of those luddite bozos who answers every call and has to deal with hanging up on the occasional spam call. :-)
* The other reasons I’m dropping Google Voice are 1) it seems like it noticeably affects call quality, at least in my area (particularly w.r.t. latency; I imagine there are fundamental round tripping issues since I’m in the midwest, and I’m sure the cell networks also have a lot to answer for here compounding the problem with the inherent latency of modern voice calls compared to the old POTS) and 2) I’m trying to get ahead of Google pulling the plug on the service and leaving me to track down and update phone number info everywhere and 3) not being able to do iOS stuff with other “blue text” phones (“call me on my other number in order to Facetime”)
My Australian mobile phone receives maybe one spam call per year. It’s my single phone and I give out the number to various businesses and people.
Also idk about other states, but in FL everything is public record, so after I bought my house I've gotten spam letters and calls like crazy.
1. Spam call from new 03 8xxx xxxx range, I was sure these were VoIP numbers handed out by MyNetFone but haven't been able to prove anything. I can't remember what the content of the call was
2. International spam call from Mozambique (?). The call never connects if I pick up.
3. Private number, recorded voice in Mandarin. I think this is a tax scam.
I'm sure I'd seen some changes in the laws regarding caller ID this year, but I can't find it.
Yes, it's much, much worse in America.
There's a national "do not call" list you can sign up to as well. This is amusing called the "Robinson List", presumably after Robinson Crusoe. The same concept is used in a dozen other countries.
I guess the penalties are severe enough that it generally only very shady people call, such as those pretending to be from Microsoft and calling to fix your computer.
The most prominent type, and have appeared in the last two years are the one-ring-only calls from some foreign country. They're trying to get you to call back to some premium-rate number.
The second are SMS based random spam. Just an hour ago I got spam from "JbHiFi" (even though it's not them, I've never given JB my number) with phishing/spam "you won a contest". I also get random SMS spam from other sources.
Then again, if you have any kind of landline in a regional area you're going to get a ton of fake NBN/Telstra/ATO/etc scammers.
All phone numbers in my Contacts app go into a whitelist that is maintained only locally on my phone.
When a call comes in, a notification gets fired off to the Firewall app, and it checks the number against the whitelist. If the number is on the whitelist, then it actually uses a functional replica of the iPhone Phone app UI to "ring" the call through.
If the incoming call is from a number not on my whitelist, then it checks the blacklist of known bad actors. If the caller is on the blacklist, then the call just gets blocked. I get a notification from Firewall that a call from a known bad actor was blocked.
If the incoming call is not on the whitelist or the blacklist, then it goes to voicemail, where it gets transcribed, and I get a notice that Firewall sent a call to voicemail.
I'm hoping that this is basically the same workflow we will see with iOS 13, just built into the basic iPhone Phone app. Then I should be able to stop using Firewall.
But I certainly don't need Nomorobo or any other such tool. Firewall solves all those problems, and does so better than any of the alternatives.
How does this one work and who provides this number ?
I ask because a twilio number is not a mobile number and therefore cannot receive 2FA from "short codes".
I am sure you know what I am talking about and have run into this ... so I ask, if you have a number that you give out to "companies" (as I also do) do you have a mobile number that can accept SMS from shortcodes ?
Sometimes sites want to know my mobile provider (why does it matter?), so I just choose Verizon or Other.
The only site that won't accept my GV number is att.com. Oh the irony...
Endless confusing configuration forms with vague repetitive labels.
The push notification necessitates the app sharing SIP credentials with Counterpath (the company that makes Bria), so it's good to have some fraud protection for the extension in Asterisk in the event that Counterpath gets breached. They are a publicly traded company, so hopefully they have decent security.
It's not useful over the open Internet but on a local WLAN or over a VPN call should work fine. Unless you use a damaged phone book storing SIP URLs for your regular contacts should work too. No messaging or fancy codecs though.
SIP calls through Twilio are $1/month plus $0.004 per minute for incoming calls. The spam scoring APIs are "costly", and maybe cost $0.01 combined per lookup.
Google Speech Recognition and TTS are virtually free for my use case. Asterisk can easily run on a $5/month VPS.
The best strategy -- get an area code that's not local and where you don't know anyone. That's what I have on my personal cell phone. Any call that comes in from that area code I know is spam. I don't answer it.
Everyone else can just hang up.
However American Telco regulations are a huge barrier to entry. There are various state and federal rules, so it wouldn't be an easy side project.
www.openphone.co
Happy to answer any questions and talk about my own setup!
I've been praying for a Google Voice replacement for around a decade now, I guess this it it!
1. Companies like Twilio largely created the explosion in spam calls by providing extremely cheap, programmable calls, so I'm always a little skeptical when the company selling the poison is also selling the antidote. 2. I really don't like the idea of a single company being responsible for a global verification system for something like phone calls. How are decisions made over what is verified (e.g. the example shows "call purpose" snippets on the screen - if I put something there saying "I have an important flight update", but then my call is "You're only getting free peanuts on your flight, but if you enroll in this MileagePlus card you'll get snack mix too!" who judges that?)
Overall I'd just say with all the consolidation of power with US tech giants I am extremely wary of giving them anything else that could consolidate that power further. I would much prefer an open, federated model.
Nah, Twilio has existed for a decade. Spam calls exploded because of court decisions https://www.apnews.com/f9e6e715ebef4c058499fef9efece977
I really hope Twilio is successful with this - it would definitely help, so long as there's a way to avoid abuse.
Just in case someone asks why I still have a "landline": it's VoIP, and I maintain it for basically two reasons:
* We use our VoIP number anytime we have to give a number out as part of signing up for something, rather than cell numbers
* It typically costs me under $25/year, which is actually mostly the 911 fee (oh, and it's an extra way to call 911, if we need to)
Yes! I did "Captcha for phone calls" for a school project earlier this year. It would play a quick message along the lines of "Hey, to ring my phone just enter these numbers: XYZ", and redirect the call to my phone if the captcha was correct.
The challenge was redirecting calls to the challenge even if they were originally directed to my phone number - the solution was to write a custom call handler (Android) that rejected calls if they weren't in my phone book, and then set up call-forwarding on rejected calls to go to a Twilio number with the challenge. There was a nice edge case where if I actually missed the call, it would just redirect you to the challenge again... but it was good enough for the class.
While testing, I accidentally actually got someone to use it! A guy repairing my watch called me, got directed to the challenge, and actually followed directions! Which was surprising to me. I also caught some obvious spam callers (same first 6 digits), and unsure if there were any other real people who did call and got confused or just hung up.
I've been meaning to clean up the code and publish it, but haven't gotten around to it - there's lots of hardcoded numbers in there.
(Disclosure: currently employed by Twilio)
It completely weeds out robocalls, and no actual humans seem to have an issue with it. On the contrary, everyone sees why it's necessary.
It has a whitelist/blacklist for common recorded messages like dentists, and then a CAPTCHA pretty much exactly what you describe: https://nomorobo.zendesk.com/hc/en-us/articles/205761825-Wha...
So, that makes solutions like Nomorobo not work so well for me.
https://support.google.com/phoneapp/answer/9118387?hl=en works on several Android/Android One phones
SHAKEN/STIR[1] is a much more network-agnostic solution to the same problem, using essentially the same, well-established technology used by internet certificate authorities. As such, it has all the same basic problems that CAs do, but it's a hell of a lot better than tightly coupling yourself to Twilio.
[1] https://en.m.wikipedia.org/wiki/Adblock_Plus#Controversies
[2] https://www.consumerreports.org/robocalls/spoofed-robocalls-...
Why would you trust Verizon/Comcast more than Twilio?
As not a customer of both, STIR/SHAKEN doesn't help me.
2. This is being proposed as part of legislation. All telecoms in the US would be required to implement the technology.
3. I don't think it's accurate to call it "open", but it's not centralized in the same way that Twilio's solution is. It's a federated system analogous to HTTPS certificate authorities. Twilio's service is more analogous to content filtering services like NetNanny. As I mentioned, CAs have problems too, but they're a whole lot more open than NetNanny.
The problem is that determining the source is difficult and even if you did it is most likely over seas where enforcing the fine is impossible.
I would love to hear a technical solution that has been implemented somewhere that had a significant impact.
> Verified By Twilio displays company branding and a reason for the call right in front of the consumer
This should make the purpose of debt collection calls obvious, though I don't see how they plan to enforce accurate call descriptions.
A reporting/flagging system, perhaps? "This call wasn't as described". After enough reports, they lose the verified status.
Typically, they do not handle their own phone numbers, their vendor for the dialer does. This is entirely dependent on the priorities of the vendor.
Due to the FDCPA, we are unable to give out the reason for our call as that would inform any 3rd party that the person we are calling to speak to is in debt.
The agents on the front line are scripted to advise them that we are calling in regards to a personal matter, or in response to the letter they received.
This gets even worse if your company name includes "Collections" or any variation of that, as they cannot give out their company name in that case unless prompted (UDAAP requires them to answer direct questions not about the loan in a straight forward way).
Whenever you receive a call you know beforehand if it's spam or not. Maybe a handful of times in the last year I've received a spam call that wasn't marked as spam. You can block callers and report as spam much like in Gmail.
When the call is not spam Android looks up on Google where the number is from and it shows the name of the business in the call screen.
This is working great for me. A year ago I received 2-3 spam calls per day which of course I've been blocking. Now I receive 1-2 spam calls per week at the most. Some weeks I don't receive any.
Edit: My phone is from Samsung (who like to add a bunch of their own software to base Android) and my carrier is AT&T.
1: https://www.theverge.com/2018/10/9/17955274/google-pixel-3-s... (I have Pixel 2, so the feature isn't just > 3 at this point)
1. How are they partnering with all of the device companies to get this on at least most phones? That'd be interesting to know.
2. I guess somebody has to type that intro text. As long as they're doing that, why bother with a normal call? How about just sending the text instead. Or send a text and let me call back for more details when it suits me, either right away or in 10 minutes. It sounds like whoever is sending these probably has a call center or somebody who's job is to sit around and answer the phone when it rings.
They are partnering with the existing spam filtering apps that users have already installed on Android and iOS. No need to involve the vendors or carriers.
2. The demo on the keynote gave a great use case where a school nurse called the speaker about their child, and she was very lucky to pick up and not let it go to voicemail.
There are still a lot of use cases for calls to hammer out important details, even if they are becoming rarer, like health or car maintenance, or places where sms is associated more with spam. This is a great way to get the best of both worlds, even if it's technically during the decline of calls.
a) if you're a sloppy enough parent that you don't even have your kids' school phone number saved as a contact, I suspect you're accustomed to relying on luck.
b) "Use our service or your children might die" is fairly over the top marketing, even for marketers.
Laws might stop some of the companies from working with the spammer, but the spammers will still find a way to spam.
A message to my doctor, bank, kids' school, etc: if you call me on a private/unknown number and don't leave a voicemail (because I can assure you that I am in that 70%) I will assume that you are spam. You don't need Twilio for this, just some basic phone etiquette.
Of course, when she made calls back with her desk or cell phone, Caller ID would display a number that was different from the main number of the clinic, and most patients would’t answer, even though they were expecting a call they’d requested. I finally built her an app that would dial the phone and spoof the Caller ID with the clinic’s main number. (This was also helpful because my wife was not the only on-call doctor, so a direct line to her was only occasionally the correct number to dial after-hours. The best move for patients was to always dial the main office number.)
I don't understand how this helps with anonymous callers?
Most corporations call from a featureless pile of numbers. Now you wont miss the important calls from numbers you'd otherwise consider anon spam.
Are that those shady apps which require your address book access just to start, aren't they?
Of course they could still be doing shady things with people's contact lists beyond that...
Instead they are actively implementing features to disclose everything to everyone.
That's bullshit. I have these important number in my address book. In the unlikely case the doctor wants to call me from a phone booth or from their friend's phone (why?), they'll leave a message and I'll call them back. As for the bank, they know my e-mail and I asked them not to use the phone for communication anyway.
Really, for many people there is no reason to answer unknown calls at all, especially if their experience is that most of these is someone wanting to sell something to them. Not to mention in our times asynchronous communication is very important for mental health.
- I don't have the 10+ numbers that dial out from my doctor's office in my phone
- I don't have the 5+ numbers that dial out from my child's school in my phone
- I don't have the 10+ numbers that dial out from my child's doctor's office in my phone
- Banks (god knows how many numbers each from multiple banks) - etc
It's literally impossible to have all the numbers I want to allow to reach me in my phone.
The fact that you happen to have a very limited set of numbers you need to be able to be called from does not invalidate the fact that many people have a vast amount of such numbers. Don't assume your anecdotal data speaks for everyone.
The concerning part is 30% still answer the calls.
A certain person who needed to give me an appointment kept calling at unpredictable afternoon and evening times from an anonymous number, ignoring the voicemails I left with my availability info, and not proposing specific times in her voicemails to me.
I did get the appointment in the end, but only by happening to notice one of her calls in real time - still from an anonymous number - and answering.
I'm told similar important uses of anonymous numbers aren't too rare in this public healthcare system.
Probably not very many. Think about it: the spammer has to pay for each minute of call time on Twilio. Let's say a spammer spams for 12 hours a day at $0.013 a minute. On average that's $9.36 a day. So if they want to do this with 24 "lines", that's $224.46 a day.
Now consider the the cost of (on-premise) phone lines provided by a company like Time Warner Cable: $25-$45 per line per month (the last time I checked). That's more like $36 a day. So with Twilio the spammer would be paying 6x the cost.
Open source is preferred, but any app with a solid privacy policy would do. It would be nice to see suggestions for both Android and iOS.
Arguably this should be administered by some trusted third party, like EFF.
[1] https://en.m.wikipedia.org/wiki/Adblock_Plus#Controversies
Hi aithrow,
One of my special interests is Whitebox machine learning, and has been for many years
With the dominance of Blackbox techniques causing lots of ethical questions to be raised on the uses of AI, I think it's time to revisit Whitebox techniques.
I’ve written a couple of non-technical articles on the subject on Medium:
Why, if you're planning to use AI you need a Whitebox system
Why are tech firms ignoring half of AI?
If you want to read about our Whitebox solutions please look at darl.ai
Whitebox machine learning is available (free) through our GraphQL API at darl.dev
Thanks for reading,
Andy Edmonds Doctor Andy’s IP sales@darl.ai
P.S. A glitch in our mail handling software meant that a small number of unsubscribes were missed on our last mailshot. If you were one of these my apologies. Just unsubscribe again, we won’t miss you this time!
Sent by Dr Andy’s Back Office AI system
If you'd like to unsubscribe and stop receiving these emails click here .
Love that Twilio is trying to solve it with a cool, seemingly useful approach.
Hate that Twilio will probably use something proprietary for this that locks people into their ecosystem even more.
Spam calls are no longer a bother Nor are really seen and important messages from those not in my contacts leave voicemails.
I don’t have their corp numbers saved and if we can see who is calling then it could be helpful.
We just started testing texting people before we call them and getting consent on an opt in form to receive a text message.
Im still not sure on the “how”
Now realize that it's not you who knows it, but rather a company that is tracking all your phonecalls and decides for you whether you should answer or not.
>Through the programmability of the Twilio platform, businesses will also be able to assign a purpose for each call to give further context.
So a malicious actor can assign a legitimate purpose to his phone call ? No thanks
Could you please, for the love of God, give us an email verb in twiml ? Please ?
