Evidence of meeting #100 for Access to Information, Privacy and Ethics in the 42nd Parliament, 1st Session. (The original version is on Parliament’s site, as are the minutes.) The winning word was platform.
A video is available from Parliament.
9:30 a.m.
Conservative
Jacques Gourde Conservative Lévis—Lotbinière, QC
You have answered my question. Thank you.
Mr. Chan, on another matter, you said just now that politicians use Facebook. We use Facebook as a means of communication, but are there political parties in Canada that advertise on Facebook?
9:30 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
I am sorry, I did not hear the last sentence very well.
9:30 a.m.
Conservative
Jacques Gourde Conservative Lévis—Lotbinière, QC
As politicians, we use Facebook as a means of communication because it is very effective. But have any political parties in Canada previously bought advertising on Facebook?
9:30 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
That do not use Facebook?
9:30 a.m.
Conservative
9:35 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
Okay.
I do not know exactly, but I believe that it is possible that each federal party represented in the House of Commons has bought advertising on Facebook.
9:35 a.m.
Conservative
Jacques Gourde Conservative Lévis—Lotbinière, QC
So it is open. If a political political party wants to buy advertising, Facebook will sell it to them. Is that correct?
9:35 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
Yes.
9:35 a.m.
Conservative
Jacques Gourde Conservative Lévis—Lotbinière, QC
If a political party buys advertising on Facebook and if people click on the party's advertising message, can that party buy more advertising later that this time would encourage people to go and vote during an election campaign, for example?
The people who click on the advertisements, whether for the Liberal Party, the NDP or the Conservative Party, are they categorized in any way? Can people receive different advertising depending on whether it comes from the advertiser or from clients?
9:35 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
Our ad products have some functionality. If the individual, the party, or the organization chooses to, they're able to ask—for those who have expressed an interest in my advertisement—if they can reach that same audience again. To be very clear, you will never know who these individuals are. So you can never go back and say specifically you want to reach Mr. Sherman, but you can say for people who commented on my post or liked my post, I would like to reach that same audience again. You have that ability if you choose to avail yourself of it. Yes.
9:35 a.m.
Conservative
Jacques Gourde Conservative Lévis—Lotbinière, QC
Mr. Chan, I understand you cannot target people, but the political party did not necessarily ask for that. The political party wants to reach those who have clicked on a party's advertising. Does Facebook directly offer a service that involves targeting those who might have an interest in a political party in particular, or is that done by a third party?
That can change everything. People say that Facebook has changed the world of information and the world of advertising, but the platform is actually also in the process of changing history. It really can influence a critical mass of people, which can change the outcome of an election campaign that might be very close. Facebook is now an integral part of our society, a social network that is influencing people's life choices. You bear a heavy responsibility. If you provide this service, it must be provided equally to everyone. If some developers have found a way to go further than the parties can, it can greatly influence the result of a vote and influence the general direction in which society is going.
How will you be able to be on the lookout for that or to provide equal services to everyone? If developers are able to open a window to get into Facebook and plug in their app, it means that they have access to a host of data that could be distributed to a third party and not to everyone. How will you go about protecting us from that.?
9:35 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
Well, sir, just to be very clear—
9:35 a.m.
Conservative
9:35 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
Oh, I'm sorry.
With respect to that, we treat all advertisers on Facebook equally, and they have access to the same products and the same services. I think I'll leave it at that.
9:35 a.m.
Conservative
9:35 a.m.
Liberal
Raj Saini Liberal Kitchener Centre, ON
Gentlemen, good morning to you both.
The other committee I sit on is the committee on foreign affairs and international development and, in January 2017, I had the opportunity to visit Latvia ahead of our troops being deployed there. One of the things we were briefed on was the disinformation campaign that would emanate prior to our troops going there, which turned out to be true.
There were many accusations levelled against our troops: that the entire deployment was gay; that they were working to turn Latvian kids gay; that they were training neo-Nazis. For me, that's a concern, because it puts the troops at risk. This reminds me of the pizzagate scandal that happened in the United States.
What are you doing to stop the spread of this information on your platforms, and how quickly are you able to react to that kind of misinformation, especially when lives are on the line?
9:35 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
Thank you very much, sir, for that question.
Obviously, I think, just to roll back a bit, looking at what happened in the U.S. presidential election, we were clearly slow to react to this. We were slow to get on top of it. I want to assure you that we're now putting in all of our efforts to address this challenge head on.
I'm not familiar with the particular example you gave. But if I may, I can give you in general terms how we think about the challenge of misinformation. It turns out, upon study and research on this phenomenon, there are two things that we've identified. One is the sort of classic clickbait, low-quality content misinformation. People may not have a particular political objective, but they're going to put stuff online; they're going to try to put stuff on Facebook. The intent is to have people click to a site where they're publishing very low-quality, potentially fake information, and get people to click through until then they monetize.
A lot of this turns out to be financially motivated. What we're trying to do, using new technologies like machine learning, the artificial intelligence that we talked about earlier, is to identify this kind of behaviour, and through our signals being able to prevent them from using Facebook ads, so effectively drying up the financial incentive to cause mischief.
9:40 a.m.
Liberal
Raj Saini Liberal Kitchener Centre, ON
I can appreciate the position that Facebook is in, with 2.2 billion users and billions of pieces of information moving on a daily basis, and I agree with that. But this was not low quality, and this affects our troops who are deployed around the world. I don't think it was done to monetize. I think it was done to spread misinformation and cause harm.
I think you're talking about two different levels. The low quality is monetization—that's fine—but there's a higher quality when you're talking about troops or other entities who are serving in different parts of the world.
I can appreciate your position because of the amount of information that's moving, but still that information has to be removed, because lives are on the line. What assurances can you give, not only to this committee but to Canadians, that troops are being protected around the world from this disinformation?
9:40 a.m.
Global Directeur and Head of Public Policy, Facebook Canada, Facebook Inc.
Absolutely, we take that very seriously, sir. Certainly trying to protect people and prevent real world harm is obviously paramount to what we do on our service.
On specific cases like that, we have a set of community standards that I think makes very clear that things we can all agree on should not be on the service, so no hate speech, no incitement to violence, no pornography, no terrorist content.
With what you're talking about—again, I'm not familiar with the specific instance—if this were reported to us and found to be in violation of these community standards, and it would appear from what you're saying that this was in fact the case, we would take it down. We have tens of thousands of people working on safety around the world, and a good chunk of those people focus on precisely what you're talking about.
9:40 a.m.
Liberal
Raj Saini Liberal Kitchener Centre, ON
It's currently still up, but I want to move to another point. I don't have that much time.
BBC recently reported that Facebook is requesting that the Canadian–European users grant them permission to use facial recognition software to identify them in photos and videos. This is an opt-in, not an opt-out service.
Given recent activities, how can we have the confidence that this new data will be handled properly? Part of this issue is to also create new friend suggestions. How exactly will it work? What will be the primer surrounding this?
9:40 a.m.
Deputy Chief Privacy Officer, Facebook Inc.
As you point out, face recognition is a feature that we're rolling out in Canada. It's something we've offered in many parts of the world for quite some time—probably around six years at this point. The primary use of face recognition technology is to suggest that people tag each other in photos. For example, if I upload a photo of Kevin, I might get a suggestion to tag him. That enables him to know that the photo exists, take action if he wants to do that, report it to Facebook if he has a concern, and all those things. We've also expanded our use of face recognition to enable people to better manage their identities so that, for example, you know if somebody's impersonating you, you know if somebody has posted a photo of you and they haven't tagged you, and also for accessibility purposes.
9:40 a.m.
Liberal
Raj Saini Liberal Kitchener Centre, ON
Because this is an opt-in feature, the GDPR has tried to limit pre-ticked boxes for opt-in consent. In this case you would have to tick the box to say you don't want to be part of it, but then you'd have to go into the managed settings to reconfirm that. Is that not a bit much? If somebody consents or doesn't consent, why do they have to go through the second step to confirm that? Why is the first step not sufficient? A lot of people may not know they have to take the second step to confirm what they wanted to do in the first step.
9:40 a.m.
Deputy Chief Privacy Officer, Facebook Inc.
We think it's important to be clear with people about how we use technologies as a part of Facebook, including face recognition. Our plan as we roll this out—
9:40 a.m.
Liberal
9:40 a.m.
Conservative
The Chair Conservative Bob Zimmer
Your time is well past, so just quickly finish your answer, Mr. Sherman.