Evidence of meeting #153 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 facebook.
A video is available from Parliament.
1:25 p.m.
Global Director, Information Policy, Google LLC
I'm not familiar with the particular cases you're talking about, but we'd be happy to follow up.
1:25 p.m.
Senator
Antares Guadalupe Vázquez Alatorre
[Delegate spoke in Spanish, interpreted as follows:]
In any case, what's next?
1:25 p.m.
Global Director, Information Policy, Google LLC
In general, if something were to violate someone's privacy or be defamatory or incite violence, and so on and so forth, against our guidelines, we would take it down. The case you're describing is something I'm not familiar with, but we'd be happy to receive more information and take it back to our teams.
1:25 p.m.
Conservative
The Chair Conservative Bob Zimmer
We'd better move on.
We'll go to Mr. Angus for just a couple of minutes, and then Mr. Collins and me.
Go ahead, Mr. Angus.
1:25 p.m.
NDP
Charlie Angus NDP Timmins—James Bay, ON
Thank you, Mr. Chair.
I want to make a confession. I'm a recovering digital utopian. I came here as a young democratic socialist and I fought hard against regulation. Imagine that, because we saw all the start-ups and we saw a great digital future. That was 2006. Now in 2019, I have conservative chairs here who are pushing for government regulation. That's the world we're in with you folks.
It's because we're talking about democratic rights of citizens, re-establishing the rights of citizens within the realm that you control. We're talking about the power of these platforms to up-end our democratic systems around the world, which is unprecedented. We're talking about the power of these platforms to self-radicalize people in every one of our constituencies, which has led to mass murder around the world. These are serious issues. We are just beginning to confront the issues of AI and facial recognition technologies and what that will mean for our citizens.
It's what our Privacy Commissioner has called the right of citizens to live free of surveillance, which goes to the heart of the business model, particularly of Facebook and Google, and it came up yesterday and today from some of the best experts in the world that the front line of this fight over the public spaces and the private lives of citizens will be fought in the city of Toronto with the Google project.
Mr. McKay, we asked you questions on Sidewalk Labs before, but you said you didn't speak for Sidewalk Labs, that it was somehow a different company.
Mr. Slater, we had experts say this is a threat to the rights of our citizens. Mr. McNamee said he wouldn't let Google within 100 miles of Toronto.
How is it that the citizens of our country should trust this business model to decide the development of some of the best urban lands in our biggest city?
1:25 p.m.
Global Director, Information Policy, Google LLC
I, too, do not work for Sidewalk Labs. You're right. We want your trust, but we have to earn your trust, through transparency, through development of best practices with you and accountability. I think then different sites will make different choices. That is in general the case, but I can't speak to that specific company because I'm not a part of it.
1:25 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
Thank you.
I would just like to pick up on a couple of things that were touched on in this session.
Mr. Potts, you mentioned briefly the changes to the Facebook Live policy as a result of the Christchurch attack. I understand that as a restriction on people who have broadcast the most serious footage through Facebook Live, and who would then have their accounts automatically suspended. Is that correct?
1:25 p.m.
Global Policy Director, Facebook Inc.
That part is correct, but also, I think, more broadly, if you have a community standards violation for specific types of actions, you would lose your access for a period of time—30 days, 60 days, 90 days—to enable yourself to go on the live product.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
You wouldn't be able to use the live product.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
Would that be at a maximum of 90 days suspension?
1:30 p.m.
Global Policy Director, Facebook Inc.
I think for egregious violations of our community standards, we also reserve the right to disable your account.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
Okay.
What about people who have shared serious content that has been broadcast through Facebook Live? They're not the broadcaster themselves, but they've shared that footage with others on the platform. Is there any action taken against them?
1:30 p.m.
Global Policy Director, Facebook Inc.
We try to look at the intent of the sharer. I think that in the Christchurch example, we had many people just sharing for awareness purposes. We had some, definitely, who were sharing for malicious purposes, to subvert our policies, to subvert our AI. Those would be actioned against. If we knew that you were sharing—even media companies shared the video—we would try to read in some context and try to understand the intent of why you were sharing.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
Did people who you believe maliciously shared the footage have their accounts cancelled?
1:30 p.m.
Global Policy Director, Facebook Inc.
In some cases there have been individuals who had their accounts disabled. In other cases they received penalties.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
Would you be able to write to us to tell us how many accounts have been disabled as a consequence of this?
1:30 p.m.
Global Policy Director, Facebook Inc.
I'd be happy to follow up after the hearing.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
These changes obviously take retrospective action against people who have done it. Is there anything Facebook has done to stop something like Christchurch from happening again, in terms of the way it is broadcast and shared through your systems?
1:30 p.m.
Global Policy Director, Facebook Inc.
We are continuing to invest in the AI.
In the Christchurch case, the use of the first-person video from the GoPro is very difficult for AI to recognize. We're continuing to invest to try to get better, to try to give training data to the machine learning so that we can identify and prevent. We have introduced new protocols for routing those types of videos to human reviewers in the moment, but it's important to note that the actual video was never reported while it was live.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
Okay. I'm not sure I'm clear on that, but there are a couple of more things.
You've said quite a lot about the deletion of inauthentic accounts. Facebook, I believe, said that there were 3.3 billion inauthentic accounts deleted over the previous six months. That is considerably greater than the active user base of the company. Based on that, how confident can you be that there are only about 5% of accounts at any one time that are inauthentic?
1:30 p.m.
Global Policy Director, Facebook Inc.
We have data science teams that study this closely, so I defer to their expertise and analysis on that.
1:30 p.m.
Chair, Digital, Culture, Media and Sport Committee, United Kingdom House of Commons
Monika Bickert said that the inauthentic accounts are far more likely to be sharing disinformation, so of those 3.3 billion accounts, how many of those were actively sharing disinformation?
1:30 p.m.
Global Policy Director, Facebook Inc.
I do not have that figure. I believe she said that they are more likely to...a combination of abusive behaviour, so not only disinformation but hate speech. Your point is taken and I can follow up.