Ryan Mac🙃 Profile picture
senior tech reporter @BuzzFeedNews. ryan.mac@buzzfeed.com & rmac18@protonmail.com. DMs open. On Signal--ask for my number
eDo Profile picture Shakeeb Ahmed Profile picture 2 added to My Authors
14 Apr
A Facebook content moderator contracted through Accenture in Austin quit this morning with a blistering internal note. I’ve typed it out in full so people can read it for themselves.
“...the current system is bad not only for content analysts but for Facebook itself.”
“I hope you figure out a way to stop constantly starting PR fires and traumatizing people en masse.”
Read 8 tweets
13 Apr
Facebook just took down a page for a QAnon-affiliated PAC called "Disarm The Deep State," after researcher @3r1nG found it and I asked the company about it last night. This comes after Facebook introduced a policy in Aug. 2020 that said it would take down QAnon pages and groups.
The page, which is associated with an FEC-registered PAC, ran targeted video ads in Florida and Indiana for candidates in local races. These ads came before Facebook's Aug. 19 QAnon enforcement, but they show the company from a QAnon-affiliated PAC.
The page wasn't hiding its affiliations. It lists a treasurer named "James Watkins" and while I couldn't confirm that was 8Chan operator Jim Watkins, the email listed is for Benjamin Barr, a known attorney for Watkins.
Read 7 tweets
12 Apr
If you're a student, you may want to be aware that campus cops or officials have used a facial recognition tool called Clearview AI. Here's a list of universities and colleges that had employees who used or tried Clearview, based on internal data (thread):
Jackson State University
University of Alabama at Birmingham
University of Alabama
University of West Alabama
University of California San Diego
UC San Diego
Southern Connecticut State University
Florida Atlantic University
Florida International University
University of Florida
University of South Florida
Columbus State University
University of North Georgia
University of Illinois at Chicago
Indiana University
Louisiana State University
University of Minnesota
Montclair State University
CUNY Brooklyn
University of Toledo Medical Center
Read 7 tweets
10 Apr
Just rented Romeo Must Die to pay respects to our king and realizing just now that it’s an amazing homage to the Bay Area
In the opening scene, these dudes drive from Marin, for some reason joy ride down Market to Embarcadero and then back thru the Tenderloin, and then they cross the Bay Bridge to go to Oakland. Just wanting to see all the sights before they break out into a needless fight smh
Just realizin how problematic this movie is in the context of current race relations
Read 5 tweets
7 Apr
Just saw this explanation from Facebook about the data leak, which interestingly claims it's focused on protecting people because "scraping data" is against its terms of service.

That's funny because FB has done nothing about Clearview AI scraping photos.
Over the last year, I have asked Facebook more than a dozen times if it will take legal action against Clearview AI for scraping what is likely millions of photos from Instagram and Facebook. No lawsuits have been filed and FB has said nothing on record.
Facebook board member Peter Thiel is an investor in Clearview AI. How can a board member, who has a fiduciary duty to Facebook, maintain a financial interest in another company that is directly harming Facebook's users?
Read 4 tweets
6 Apr
We spent a year reaching out to (and following up with) all the taxpayer-funded organizations listed in the data, so it's extremely rewarding to see people use our table to see if local police or government agencies in their area used Clearview AI.
Even with our reach outs, the majority of organizations in the dataset did not respond to our questions. But the reporting shouldn't end there. With the table, we're hoping that local reporters and concerned citizens can continue to hold these public servants accountable.
Of the 335 organizations that confirmed their employees used or tried Clearview, 69 initially denied any use before we questioned them further, leading them to uncover officers or individuals that had used free trials. That was a big impetus for our story. buzzfeednews.com/article/ryanma…
Read 5 tweets
6 Apr
It took more than a year to report this story. We contacted more than 1,800 US taxpayer-funded organizations listed in data as having using Clearview AI to ask about facial recognition. Here's a thread of some our most important findings:
Of the 1,803 entities listed in the data as having run at least one search with Clearview AI by Feb. 2020, 335 orgs confirmed to BuzzFeed News that they had used or tried the facial recognition software. buzzfeednews.com/article/ryanma…
The 1,803 entities listed in the data had run almost 340,000 facial recognition scans from 2018 to Feb 2020. The largest searchers, according to the data, included the NYPD, FBI, CBP, and ICE.

Read 10 tweets
30 Mar
Lol they wrote up a whole story about our scoop and couldn't even bother to credit, much less link wsj.com/articles/faceb…
gee, i wonder what made them have to confirm it in the first place? Image
Big pubs may not credit other pubs because they don't want to make it appear that they're second to a story, but it's a fundamental disservice to readers. Hiding information is never a good look. Crediting doesn't make your story any worse.

Read 4 tweets
25 Mar
God this is gonna suck
These two opening statements make it pretty clear we're going to get nowhere. One side cares about the spread of misinformation. The other cares about "woke books" and calls the removal of misinformation "censorship."

Good luck to everyone watching this for the next 6 hours.
That talking point from Mark Zuckerberg about Facebook not being responsible for polarization? That came straight out of a presentation Facebook made to employees earlier this month. buzzfeednews.com/article/ryanma…
Read 11 tweets
18 Mar
Facebook is building an Instagram for children under the age of 13, based on internal docs and interview with @mosseri. My latest with @CraigSilverman.

On Tuesday Facebook announced that it was trying to make Instagram safer for teens. Nowhere in that announcement did it say that it was building a product for children under 13. We got that internal announcement, which went up today. buzzfeednews.com/article/ryanma…
After seeing the internal docs, I asked @mosseri about it. He said there is no "detailed plan." The thinking was that kids want to join Instagram anyway and that FB wanted to build a product with safeguards and age verification with tracking for parents. buzzfeednews.com/article/ryanma…
Read 5 tweets
17 Mar
Sure this is important, but 8 people were just killed in a shooting in your state and you’ve tweeted nothing about it?
The pinned tweet is just the icing Image
His office has said nothing as well. It’s not something that’s hard to denounce. Image
Read 4 tweets
12 Mar
In a webinar for employees yesterday, Facebook execs presented talking points so that their workers could explain to friends and family why the social network doesn't cause harmful polarization. buzzfeednews.com/article/ryanma…
The presentation came with an internal "one-pager" that included academic research (some of it funded by Facebook). At one point, an employee asked if the company was like Big Tobacco funding research about its own products to show they weren't harmful. buzzfeednews.com/article/ryanma…
Key in all this is that FB has teams dedicated to studying this stuff internally. But none of those findings were presented in this "one-pager." One former employee who studied polarization told me that the document amounted to "corporate gaslighting." buzzfeednews.com/article/ryanma…
Read 7 tweets
11 Mar
Facebook is holding an internal employee meeting about "Polarization And Our Products." Execs are discussing reports about how FB leads to societal polarization.

One employee in the meeting just basically asked, "How are we different than Big Tobacco funding its own research?"
The response was that Facebook was not funding its own research in this specific instance.

Facebook does fund numerous outside research projects, for example this NYU groups study that @CraigSilverman and I wrote about: buzzfeednews.com/article/craigs…
During the event, execs are referencing a document written by Chief Product Officer Chris Cox.

At the very top of that paper it reads: "TL;DR Despite media claims, there is limited causal evidence linking Facebook use with harmful 'affective' polarization."
Read 5 tweets
3 Mar
David Brooks' think tank project has been funded by Facebook and various billionaire families. Over the past two years, he's written about the project (and FB) in his NYT column, though he's never disclosed that funding and his salary for the project.
Facebook gave David Brooks' Weave Project (which is part of the Aspen Institute) $250k in Dec. 2018. That has never been reported before, not even in the Aspen Institutes' disclosure reports.
Employees at the NYT were furious about our first story, which disclosed Brooks wrote glowingly about FB for FB's corporate site last month without telling his editors. For the past few years, Brooks' NYT colleagues have reported critically on the company.
Read 6 tweets
25 Feb
At today's Facebook companywide meeting, Facebook execs discussed the possibility of adding facial recognition to its upcoming smart glasses product after one employee brought up concerns about "real world harm" including "stalkers." buzzfeednews.com/article/ryanma…
Facebook is coming off a $650 million settlement from last year for violating Illinois' Biometric Information Privacy Act.

FB VP @boztank said the law may prevent the company from deploying facial recognition on smart glasses, but noted the co is considering its options.
FB has one of the world's largest user-uploaded photo databases, and has been working on FR for years. That tech powers features like suggested tags. Deploying it on smart glasses to help ppl recognize unknown faces in the real world would be totally new. buzzfeednews.com/article/ryanma…
Read 4 tweets
25 Feb
We've seen how political considerations play a role in Facebook's domestic decision-making, but what about abroad? This from @jackgillum and @JustinElliott is such a telling look into how the company caves to those in power when threatened w/ a shutdown. propublica.org/article/sheryl…
As Americans, it's easy to forget Facebook's impact on a global scale and how its execs' smallest decisions (or non-decisions) have an outsized impact on whole countries. And it's interesting to see how PR concerns play into decision-making abroad as well. buzzfeednews.com/article/craigs…
Facebook's actions here, as laid out by the ProPublica story, provide a playbook for in-power regimes. Don't like something? Threaten to block FB in the country and kill ad revenue. FB may cave and its execs will likely be "fine with this."
Read 4 tweets
24 Feb
Is this.... car crash sponcon
The marketing folks over at Hyundai saw the iconic Axe spray can insurrection tweet and said, "Yeah, we can do better."
Alright this is weird
Read 7 tweets
24 Feb
Mark Zuckerberg said Facebook would be cutting down on political content in news feed but I just got this because a friend is arguing in the comments section. There is nothing I’ve engaged with that suggests I’d read this.
Facebook's News Feed algos control how this content is fed to users.

Remember: FB has a "hate bait" dashboard that shows that pages like Breitbart tend to attract the most divisive, hateful comments from users. Knowing that, why boost and spread it? buzzfeednews.com/article/ryanma…
And before the FrEe spEEcH crowd gets a hold of this, this has nothing to do with the content nor the ability of the page to say this. The point is to show the post's distribution.

There's no reason for me to see this and I don't want to see this, so why put it in front of me?
Read 4 tweets
24 Feb
It’s been a privilege to work alongside @CraigSilverman and I am so happy to share this with him.
This whole thing has been a team effort. From our @BuzzFeedNews editors @JohnPaczkowski @mat @SchoofsFeed to our @fwd reporting team and so many others that helped get these stories over the line in the last 8 months.
Read 4 tweets
23 Feb
Not much internal interaction on our latest Facebook story besides one manager blaming leaks and including a wartime poster photo w/ the phrase "Loose Lips Sink Ships."

"We need people to stop leaking. It's honestly that simple," he writes.


"Im not sure how our comms team is expected to help get our side of the message across as long as reporters can seemingly get everyone at the co to provide out of context quotes at will."

(I'm quoting him in full so I uhhh can't be accused of lack of context.)
Our story was based on internal docs — quoted with plenty of context — and 14 interviews with current and former employees. We asked for interviews for execs, but they refused.

How's that for context?
Read 5 tweets
21 Feb
The latest Facebook story from @CraigSilverman and me details a history of how political concerns influence important platform decisions. While employees are warning about it now the 2020 election is over, there have been plenty of signs in the past. buzzfeednews.com/article/ryanma…
There's been a lot of amazing past reporting that we built upon. For example this story from @sheeraf @nickconfessore @ceciliakang @AllMattNYT and @jacknicas on Facebook's fallout with the 2016 election and Cambridge Analytica. nytimes.com/2018/11/14/tec…
.@dseetharaman, @JeffHorwitz, @newley and the WSJ team have led as well. Deepa's piece on Joel Kaplan and Jeff and Newley's piece on Facebook's concessions to political concerns in India are fundamental to understanding the co.


Read 7 tweets