Tech Policy Leaders

The discourse in the US is increasingly toxic and divisive. Disinformation, misinformation, and hate speech are rampant and we have few options to change it.

The US has been entrenched in a deteriorating political and information environment for a long time now. The Supreme Court's overturning of Roe v. Wade instantly removed a right that 165 million Americans expected to have indefinitely.

The discourse has always been divisive, but now it's reaching a point where people are unable to listen to one another or understand each other's perspectives. This means that there isn't any real discourse anymore, just fighting words that make people feel good about themselves but don't actually accomplish anything at all.

Bio

Ryan Merkley

@ryanmerkley

Ryan Merkley is Managing Director at Aspen Digital, focused on emerging technology, internet policy, and the information ecosystem. He is an accomplished executive working at the intersection of public good and technology for organizations like Wikimedia, Creative Commons, and Mozilla.

Resources

Commission on Information Disorder Final Report

Supreme Court overturns Roe v. Wade, opening the door to surveillance

In a 6-3 decision Friday, the Supreme Court overturned Roe v. Wade. Justices Breyer, Sotomayor, and Kagan were the only dissenting justices. Writing for the majority, Justice Alito left it up to state legislatures to write their own abortion laws. As far as tech policy is concerned, many advocates, including WashingTech, are concerned that law enforcement will now be able to surveil location data in any of the 13 states in which abortion is now outlawed.

Congress inches closer to federal privacy law

The House Energy & Commerce Committee passed a bipartisan privacy framework on Thursday, with the measure now heading to the Senate. Reuters' Diane Bartz reports the bill would let you opt out of targeted ads online. It would also give users the ability to sue firms for selling their user data to third parties. The bill would override the patchwork of state privacy laws we have currently in states including California, Colorado, Connecticut, Utah and Virginia. The Washington Post reported Thursday that Senator Maria Cantwell, a key vote, doesn't support the bill in its current form for precisely that reason – the Senator believes that, in many cases, the state privacy rules are stronger than the ones in the House bill.

Greintens may have incited violence on social with 'RINO' hunting video

Eric Greitens, the former Republican governor of Missouri who is now running for the US Senate, posted a video appearing to encourage viewers to go 'RINO' hunting – RINO being an acronym for Republican in Name Only. "Join the MAGA crew," Greitens says in the video, "Get a RINO hunting permit." He says this as he's holding a shotgun surrounded by smoke, and a couple of boneheads dressed up as SWOT team officers bust through the door behind him. Facebook was the only company to remove the video outright. Twitter and YouTube left it up, although Twitter added a "public interest notice" to the tweet.

Talk therapy apps under scrutiny

If you've used a talk therapy app like Betterhelp or Talkspace since COVID started, chances are you've had at least a fleeting concern about how these companies use your data. Well, Senators Elizabeth Warren, Cory Booker, and Ron Wyden are concerned too, which is why they sent a letter to the firms asking them to explain their privacy practices. The Democratic lawmakers want to know how these firms collect data, what they do to protect it, and how they communicate their data protection practices to their users.

Meta reportedly to shut down misinformation-reporting tool

The Verge reports that Meta is planning to shut down Crowdtangle, which can be used to find misinformation within popular social media posts. A Meta spokesperson told the Verge that the company will probably keep Crowdtangle working at least through the midterms. After that, the company says it plans to launch a better product.

Amazon may enable your Alexa assistant to take on the voice of a dead relative

So at its annual re:MARS conference in Las Vegas, Amazon SVP and Head Scientist for Alexa, Rohit Prasad, demonstrated how a future iteration would enable your Alexa Assistant to take on the voice of anyone, including a dead relative, at least that was how Prasad decided to demonstrate the product – with someone's dead grandma reading a bedtime story.

Don't ask me – I'm just reading what they wrote for me here.

Anyway, that's it for this week. You can find links to all of these stories in the show notes.

Stay safe, stay informed, have a great weekend. See you Monday.

Biden calls for better kids privacy laws in State of the Union

During his state of the Union address Tuesday, President Joe Biden called for better regulation of social media companies. First Lady Jill Biden invited Facebook whistleblower Frances Haugen to the State of the Union – Haugen was the first to shed light on Facebook's (now, Meta's) internal efforts to target children as young as 6 on Instagram, and the fact that the company ignored its own research showing Instagram damaged teenage girls' self-esteem.

Facebook is getting sensitive medical information

The Markup reported yesterday that Facebook may have been receiving your medical information from a tracking tool – Pixel. Many hospitals use Pixel on their websites to track site visits. So let's say hypothetically that you search for a health condition on the hospital's website – well, for about a third of those sites, the tracking tool sends the information to Facebook. Johns Hopkins, UCLA ReaMgan, New York Presbyterian, Northwestern Memorial, and Duke University Hospital are among the hospitals that track site visitors with Pixel.

Democrats led by Sen. Warren introduce bill that bans sale of location data

In the wake of the leaked Supreme Court decision to overturn Roe v. Wade, which paves the way for red states to criminalize abortion procedures, Sen. Elizabeth Warren led a group of Democratic lawmakers to introduce a new bill – The Health and Location Data Protection Act – that would completely ban the sale of your location data. The bill also envisions empowering the Federal Trade Commission to intervene when necessary.

Advocates warn of hate speech problems in Klobuchar's antitrust bill

Advocates including Free Press are pushing back against Amy Klobuchar's antitrust bill – the American Innovation and Choice Online Act – because they're concerned the bill would let companies that may have profited from hate speech and disinformation – like Infowars – to sue platforms like Google from banning them in search rankings. Advocates worry a provision that prohibits Google from favoring their own search results over smaller competitors – could pave the way for disinformation profiteers to make anticompetitive accusations when platforms ban their sites.

Elon Musk suggests harmful content on Twitter should stay up if it's just entertainment

Elon Musk announced to Twitter employees in a livestream that free speech issues should outweigh content moderation. Last month, said that he'd reinstate Donald Trump's account, although he said that prior to the commencement of the January 6th committee proceedings. Twitter's stock has steadily dropped from $54.20 per share, which was the price when Elon Musk made his $44 billion bid to purchase the company. At the closing bell today, Twitter was trading at $37.78.

That's it for this week. You can find links to all of these stories in the show notes.

Stay safe, stay informed, have a great weekend. See you Monday.

I decided to do a solo episode this week because I think it's really sort of super important to highlight bias in the public policy profession – because it is a profession. Because over the last 17 years that I have been working on tech and media public policymaking, majority-white organizations have always seemed to think it's totally fine to attack organizations founded and led by people of color, orgs like this one, to pursue their status ambitions. So, let's be transparent, shall we?

Just to give you some background – when I started working in this space – there was one organization in Washington – the Multicultural Media & Telecom Council (MMTC) – that focused specifically on telecommunications and media policymaking as they relate to underserved and underrepresented communities. This is where I cut my teeth as a young lawyer, in the ONLY fellowship in town for somebody like me who went to law school at night. Their members are, and continue to be, some of the finest minds in the business – folks like Ari Fitzgerald, a partner at Hogan Lovells who was on the show last week.

These are lawyers of color primarily, who are at the top of this craft.

And let me interject something here – this isn't Oakland. We are not Color of Change.

This is DC, it's a political town, it's buttoned-up, and comparing orgs of color in this town to orgs like Color of Change – is really just not a relevant comparison. A better comparison would be to an organization like NAACP or the National Urban League – both of which have local chapters but they're based in DC.

Donations come from 2 primary categories here in DC – corporations and foundations – that's it. If you've been around a long time – like the NAACP, NUL, or AARP – you have members. You can raise money from them, in addition to seeking other forms of support. That is the way this market works.

Back around 2011 and 2012 it was organizations like Public Knowledge and Free Press calling out MMTC for accepting donations from Comcast. Again, one of the worst companies in the world for customer service. But they are internet service providers. It didn't matter if larger, huge nonprofits worked with these same companies – all that mattered was that Free Press and Public Knowledge needed someone to pick on when they were advocating for net neutrality. MMTC opposed net neutrality, which didn't make sense to me, which is why I started this organization when I was laid off from the Joint Center, where I co-led an institute along with Nicol Turner-Lee, which was also focused on telecommunications and media policy at the intersection of communities of color.

So-called progressives with super-deep pockets didn't like orgs like MMTC because they were the only game in town – they had too much credibility – and they opposed net neutrality ( for reasons, by the way, I continue to be baffled by – but, in any case, they opposed it). So orgs like Public Knowledge and Free Press called them out – and I called out Public Knowledge and Free Press for having ZERO people of color working there but somehow having the audacity to try to drag MMTC.

And as a result of my advocacy – at least I like to think it was – since no one else was pushing back – Public Knowledge is led by the great Chris Lewis and Free Press, by Jessica Gonazlez, who serves co-president along with Craig Aaron.

These orgs now actively recruit diverse talent – they have changed drastically – and I'm proud of them. Jessica, Craig and Chris are my colleagues – just like you have colleagues in any profession – they fixed their model and stopped attacking MMTC. We'll see what happens when and if the net neutrality debate starts up again. But, for now, we're good. Joe Torres is at Free Press who wrote the book on diversity in the news profession.

SO LET'S forward to 2016. In 2016, when this organization – WashingTech – was still an LLC, for profit, Chanelle Hardy, someone I've known since I first moved to DC in 2005 – who had worked at the National Urban League, on the Hill, and the FCC – joined Google. And, again, as a result of my advocacy, since I was vocal about it, as I am now – one of our taglines then and now – is the Inclusive Voice of Tech Policy, since nobody else cared about that until George Floyd. In 2016, Google started a cohort of folks called Next Gen Policy Leaders – the only PROGRAM IN TOWN AT THE TIME – to engage and involve people of color in tech policy. I continue to participate in the program because it is educational, offers great networking, and, again, continues to fill a need that everyone else just woke up to a couple years ago: the lack of diversity on panels, at networking events, on faculties, you name it, related to tech policy issues.

Google was a first mover, while the rest of these tech companies, and nonprofits, were asleep.

So, whose fault is that? Whose fault is it that they've built loyalty by investing in us.

Now, fast forward to today – here comes another organization – the so-called 'Tech Transparency Project,' which, again, isn't a racially and ethnically diverse organization, attacking Google Next Gen in a mediocre paper suggesting that Google had bought out people of color so they wouldn't speak out about Timnit Gebru's firing – again Dr. Gebru is an engineer whom Google fired for speaking up about bias in one of Google's algorithms.

First, I OPPOSED her firing, and vocally, on a listserv read by many Next Gens and other people of color in this space. I blasted Google for it. I was livid. I was so vocal, no one else in the Next Gen cohort needed to be – which is always the case in this town – and let me tell you something, we get PENNIES compared to some of these larger organizations.

Do you know what Google donated to us last year? $35,000. The rest of our funding came from Foundation support. But let's take a look at the Center for Democracy & Technology (CDT), which published their annual report last week. Let's see what Google donated to CDT. And CDT is a partner of ours. I'm on their advisory council. Many of their fine scholars have been on this show. But let's take a look – How much did Google donate to CDT in 2021? Wait for it! Over $500,000. Two other donors gave that much – the Chan Zuckerberg Initiative and the Knight Foundation.

Now, let's use CDT's search feature on its website to see how much work they've done on "Timnit Gebru." How many times did CDT so much as mention Dr. Gebru's name, much less call out Google for firing her? ZERO. How many times has the Tech Transparency Project called out CDT for failing to discuss Timnit Gebru? ZERO.

So come on, let's talk about 'Tech Transparency,' everybody. I guarantee I've made more sacrifices in the cause of inclusion and so-called "transparency" in this space than most of Washington. Let's talk about it.

Congress makes progress on federal privacy law

A bipartisan group of lawmakers in the House and Senate introduced a privacy bill last week that some observers including the Washington Post say faces a steep uphill battle. But the bill would require companies that use your data to only collect that data which is necessary for their businesses to function properly. It also proposes a number of other things such as a requirement for the Federal Trade Commission to keep a database of data brokers.

Amazon may have to pay for work-from-home equipment

A federal judge denied Amazon's motion to dismiss a lawsuit brought by a California engineer who works for the company from home. The employee brought a class-action lawsuit against Amazon for failing to reimburse him for equipment and internet service required to carry out job-related duties. The case now heads to trial in California.

Family sues Meta for daughter's self-harm, eating disorder

Remember the Facebook papers? that trove of documents company whistleblower Frances Haugen released last year showing that Facebook knew it was harming the self-esteem of young girls and continued doing it anyway? Well, the family of a 19 year old young woman is now suing Meta in the Northern District of California saying the company turned her daughter from a bright and happy child into one that has engaged in self-harm and been hospitalized for depression.

YouTube proves you can train AI to spew out hate speech automatically by feeding it 4Chan posts

A YouTuber ran an experiment training an algorithm on typical 4Chan posts spewing hate speech. For those who don't know, 4Chan is a public bulletin board where anyone can make posts anonymously. 4Chan is known for its users' racist, sexist, and nihilistic posts. So YouTuber and AI researcher Yannic Kilcher took 3.3 million 4Chan threads and fed them into an algorithm. He then set the algorithm to start posting on 4Chan, and lo and behold, the algorithm produced vile posts of its own. The research is important because it suggests that any bad actor can set an algorithm to post fake or misleading information by simply using a single algorithm.

Republicans jump behind Elon Musk's bid for Twitter

Republicans jumped in to defend Elon Musk in his accusations that Twitter was attempting to thwart his $44 billion offer to buy the platform. Musk argued that Twitter was refusing to provide important documentation about bots on the platform. Texas AG Ken Paxton then launched an investigation into whether Twitter was using unlawful means to obstruct the deal. Twitter ended up granting Musk access to the information he requested, while arguing that Musk's accusations about bots was simply a pretext to back out of the deal. Since Musk officially announced his plan to purchase the platform back in May, conservatives have rejoined the platform in droves.

Feds to investigate Tesla's autopilot crashes

There's a reason why you keep see ing Tesla's bashed apart. It's because the autopilot feature isn't working properly in many of these vehicles, and it's causing the cars to crash into stationary vehicles, including police cars. That's according to the National Highway Traffic Safety Administration, which has stepped up the probe it began conducting into Tesla's autopilot feature last year.

Most of us probably don't think about it much, but our cell phones and Wi-Fi connections use something called spectrum to send and receive data. The term "5G" refers to the fifth major version of this standard for transmitting information wirelessly. It's also known as "millimeter wave technology."

The 5GHz spectrum has long been viewed as a way to get around the crowded radiofrequency (RF) spectrum and its associated problems with Wi-Fi. The widespread adoption of wireless local area networks (WLANs) in the 2.4 GHz ISM band means that there are very few available channels in any given location, which can lead to high interference levels, reduced coverage range, and low throughput.

The 5GHz spectrum is not as crowded as the 2.4GHz band, making it a better choice for high-throughput, low-interference deployments like indoor video surveillance networks. In general, this band also provides greater immunity from interference from devices like baby monitors or cordless phones that use frequencies in the 2.4GHz band.

Ari Fitzgerald

LinkedIn

Ari Fitzgerald is a Partner with the law firm of Hogan Lovells, where he leads the firm's Communications, Internet, and Media practice. He provides strategic, legal, and policy advice on a wide range of communications and spectrum policy issues to some of the world's largest and most dynamic communications network operators and equipment manufacturers, as well as industry trade associations and investors.

Resources

Hogan Lovells' Communications, Internet, and Media Practice

Sheryl Sandberg is leaving Meta (née Facebook) after 14 years with the company

Sheryl Sandberg is leaving Facebook after 14 years with the company.The Wall Street Journal reported that Meta has been reviewing Sheryl Sandberg's personal activities, including a reported allegation from the Journal's sources that she used company funds to plan her wedding. But Meta spokeswoman Caroline Nolan denied Sandberg's departure was related to the review.

The tech sector and civil rights advocates achieve small victory in Supreme Court

The tech sector and civil rights advocates succeeded before the Supreme Court after a 5-4 decision to stop HB20 - a Texas social media bill - from going into effect. The Republican bill would give citizens the right to sue platforms with over 50 million users for censoring their content for political reasons, or based on their "viewpoint." Many conservatives believe the larger platforms discriminate against conservative speech. Smaller social media companies catering to conservatives, like Parler, however, wouldn't be subject to the Texas law.

So they'd be able to censor progressive speech.

Justices Alito, Thomas, and Gorsuch wrote the dissenting opinion, which Justice Kagan didn't join. However, Justice Kagan was the 4th justice who voted to leave the Texas law in place. The Court didn't provide the rationale for their decision. WashingTech filed an amicus brief in support of overturning the Texas law.

Period-tracking apps aren't regulated

Yesterday Democratic Congresswoman from California Sara Jacobs introduced a bill called the My Body, My Data Act. The bill comes on the heels of the leaked Supreme Court opinion to overturn Roe v. Wade. In a post-Roe v. Wade world, there is a forseeable risk that period-tracking apps would send data to law enforcement in areas where abortions become illegal. \

University of Maryland: Nearly 2/3rds of Republicans want net neutrality rules

The University of Maryland published a report showing almost 2/3rds of Republicans would bring back at least some of the net neutrality rules the Trump administration overturned. Pure net neutrality rules would prevent internet service providers from discriminating against different types of traffic, prioritizing some traffic over others. Some conservative advocates want the same standard to apply to social media platforms. That means both the Twitters of the world and the internet carriers themselves could be held liable for discriminating against content, which, frankly would then have to be applied to broadcasters.

Finally, a new paper from UNC argues that Gmail's spam filter might exclude things the intended recipient actually wants to read. Conservatives jumped on the study and started saying it proves Google's alleged anti-conservative bias. But the academics who wrote the study say that interpretation takes the study out of context. Obviously, mis and disinformation and hate speech filters have a completely different design, it's a whole different set of keywords, but, sometimes you just have to listen and let people have their say.

That's it for this week! I hope you had a great Memorial Day weekend and week! Happy Summer!

And remember, stay safe & informed. Have a great weekend.

Washington Post reports on "light-sensing" technology to identify weapons

A company called Evolv uses "active sensing" — a light emission technique also used in radar and lidar—to create images and identify weapons in public and private venues, according to the company. Despite "fundamental limitations in differentiating benign objects from actual weapons," existing clients include the New York Mets, Lincoln Center in New York City, the Charlotte-Mecklenburg school system, and Six Flags amusement parks across the U.S.

Washington Post questions legitimacy of claims that social media plays outsize role in gun violence

Writing for the Washington Post's Technology 202 newsletter, Cristiano Lima analyzed the legitimacy of arguments being made by government officials like Republican Texas governor Greg Abbott, that place undue focus on social media when it comes to gun violence, detracting from the real issue which is lax gun control laws.

Activist's protest Amazon's work with police and immigration agencies

Black and Muslim activists led a protest during Amazon Web Services' summit on Tuesday. The protesters opposed Amazon's work to facilitate the surveillance and deportation of immigrants and people of color. Amazon removed three Black or Latino activists who had registered for the summit.

Digital surveillance affects 227,000 immigrants

A critical new report by NoTechforICE details how digital surveillance systems impact immigrants' lives. Constant surveillance, which is touted as an "alternative" to traditional detention, has placed more than 227,000 immigrants under some form of surveillance. These anxiety-inducing technologies interfere with employment and further stigmatize immigrants.

White House issues Executive Order to study facial recognition and predictive algorithms in the criminal justice system

The National Academy of Sciences will examine how facial recognition and predictive algorithms are being used in the criminal justice system, in order to surface civil rights issues and make recommendations to correct them. The order also requires the Attorney General to perform a disparate impact analysis of PATTERN, the Bureau of Prisons' risk assessment program , to see how it affects inmates' chances of early release.

New report details how law enforcement can use technology to enforce abortion laws [Wired]

Wired reports on a new study from the Surveillance Technology Oversight Project that details how law enforcement can utilise existing data access technologies and tracking tools to enforce abortion bans. The report cites keyword search warrants and geofence warrants as examples of these these types of surveillance technologies, which the report states can harm those looking for abortion providers or obtaining abortions.

Citing potential harm to abortion seekers and providers, lawmakers urge Google to stop the excessive retention of location data

Separately, Democratic lawmakers wrote a letter urging Google to stop its practice of retaining what the representatives believe is too much location data, which they argue law enforcement could use against those seeking abortions at abortion clinics.

New CDT report looks at surveillance of disabled people

According to a new report by the Center for Democracy & Technology, algorithms and surveillance technologies are being used to surveil, control, discipline, and punish people, with particularly harmful effects on disabled people in education, the criminal legal system, healthcare, and the workplace.

AI is an interdisciplinary field that draws on many fields of study, including computer science, psychology, neuroscience, and mathematics. AI researchers often bring a social science perspective to the field. They want to understand the social implications of AI and to identify ways that humans can best interact with AI systems. Social scientists look at how people interact with AI systems and how these interactions might change over time. Social scientists also want to understand how AI systems learn from human behavior and make predictions about how people will behave in the future.

There are two main types of social science approaches to AI. The first is an approach that focuses on the ethical implications of AI. This approach assumes that we cannot build AI without carefully considering its ethical implications for society. The second type of social science approach is an approach that looks at the psychological impact of AI on humans. This approach assumes that humans could feel uncomfortable or threatened by the presence of AI in their lives. Social scientists use these approaches to help us better understand and predict the impact of AI on society as a whole.

Emnet Tafesse and Ranjit Singh contrast the differences in how researchers in the "Global South" and "Global North" investigate the social impacts of artificial intelligence.

Bio

Emnet Tefesse

@emnetspeaks

Emnet Tafesse is a Research Analyst at Data & Society on the AI on the Ground Initiative. She has a passion for utilizing advocacy, research, and policy to create positive social change and a more equitable world. She received her Master's in Public Policy from the University of Chicago and her BA in Political Science and Sociology from Howard University.

Ranjit Singh

http://ranjitsingh.me/

Ranjit has a doctorate in Science and Technology Studies (STS) from Cornell University. His research lies at the intersection of data infrastructures, global development, and public policy. He uses methods of interview-based qualitative sociology and multi-sited ethnography in his research. He examines the everyday experiences of people subject to data-driven practices and follows the mutual shaping of their lives and their data records.

Resources

AI on the Ground Blog Series

NY Attorney General Opens Twitch Probe in Buffalo Shooting Aftermath

New York State Attorney General Letitia James opened a new probe in. to Twitch, 4chan, 8chan, and Discord in the aftermath of the mass hate-motivated shooting at a Buffalo supermarket that left 10 Black people dead. The probe is geared towards determining how the shooter used these platforms to plan, discuss, stream or promote the mass murder to which the defendant, who was apprehended at the scene of the shooting, has pled "not guilty."

Federal Trade Commission Votes to Hold EdTech Companies Accountable for Kids' Privacy

The Federal Trade Commission (FTC) voted unanimously yesterday to adopt a policy statement cracking down on how EdTech companies use kids' data beyond enabling their ability to do homework or attend class remotely. The statement warns EdTech Companies that it is against the law for these companies to require parents and schools to give up their childrens' privacy rights in order for them to access these critical apps. The FTC told EdTech companies that they should also expect fines for using kids' data for marketing purposes, keeping kids' information for longer than necessary, or not having proper safeguards in place to maintain kids' privacy while they're using these apps. President Biden commended the FTC in a statement.

Twitter Adds New Content Moderation Rules Despite Elon Musk's $44 Billion Bid

Twitter actually expanded its content moderation rules despite Elon Musk's $44 billion bid to buy the social media platform. The move came the same day Twitter GC Vijaya Gadde said "there's no such thing as a deal being on hold," after Musk's statement that he was reconsidering his hostile takeover. The new content moderation rules target misinformation about wars and natural disasters by preventing users from retweeting misinformation about things like the war in Ukraine, and preventing misinformation from appearing in search results.

Texas AG Sues Google Over 'Incognito Mode'

Texas Attorney General Ken Paxton added another allegation to the lawsuit it filed against Google back in January, along with the States of Washington, Indiana, and the District of Columbia, over Google's location settings. Paxton says that Google's incognito mode isn't actually private at all, and that it risks exposing a users political leanings or sexual orientation. Google is disputing the claims.

Instagram Removes Summer Walker's Post Criticizing KKK

Instagram left a lot of folks scratching their heads after it removed R&B singer Summer Walker's post criticizing the KKK. Walker's post simply said "why the KKK never got a RICO," RICO, of course, being the Racketeer INfluenced and Corrupt Organizations Act, which used to target the mafia, but now targets street gangs, according to VICE. Asking why the KKK never got a RICO charge seems like a damned good question, and it's interesting that Meta took that down as so much white supremacist hate speech stays up. We need more intel on who exactly within these companies is making the last call on these takedowns.

MIT Finds Medical AI Can Predict Patients' Race

A new MIT study reviewed several medical imaging technologies and found they did a better job identifying the race of patients than humans. But the researchers have no idea how this was possible. They looked at chest X-rays, chest CT scans, and mammograms and the AI identified the patients' race, even though the patients hadn't identified what their race actually was.