Securing Democracy Misinformation Disinformation and the Cyber-War for the 2022 Midterms

Show video

hello um good evening and welcome to the john f kennedy um forum my name is miss kalith and i'm a sophomore at the college studying neuroscience i'm also proud member of the jfk junior committee forum um and so just a quick few um housekeeping notes before we start so please note the exit doors which are located on the jfk street side of the forum so it's on the left and the other one on the right so in case of emergency please locate those in advance um in addition to that please also take a moment now to silence your phones um and yeah please take your seats now and join me in welcoming welcoming my fellow student sarah solomon who will introduce our guests [Applause] good evening everyone and welcome to today's john f kennedy junior forum at the institute of politics my name is sarah solomon and i'm a junior at the college studying government and serve as a member of the forum committee in the aftermath of the 2020 election amidst the turmoil of the covid pandemic the u.s witnessed a rising claims of election fraud and interference despite the cyber security and infrastructure security agencies release statement assuring this election to have been the most secure in american history this discussion continues into the upcoming midterm elections especially surrounding the use of social media as a mobilizing tool to disseminate election miss and disinformation this evening we are privileged to be joined by chris krebs the former head of the u.s cyber security and infrastructure agency during the 2020 election who has had an esteemed career of service in the department of homeland security as the under secretary for the national protection and programs directorate and the assistant secretary for infrastructure protection following his government tenure mr krebs has founded a cyber security consultancy the krebs stamos group where he currently serves his partner we are also honored to be joined by our moderator dr joan donovan who serves as the research director for the shorenstein center on media politics and public policy at the harvard kennedy school in this role dr donovan largely specializes on the consequences of technology for civil society and democracy including her trailblazing work on online extremism disinformation and social media before i turn it over to dr donovan we have one programming note for next week we hope you will join us back at the jfk junior forum for the c-more e and ruth b harris lecture given by former massachusetts governor deval patrick a graduate of the college in the law school who will share his perspectives gained through a distinguished career in public and private life thank you all very much for joining us tonight and we hope you enjoy this evening's discussion [Applause] thank you so much for the warm introduction it is really great to be here at iop again and very thankful to the shorenstein center and the belfor center for co-sponsoring i do want to warn you that i have uh i'm just gonna fix this little thing keep sliding off but uh i've stalked the crowd unfortunately in my favor and uh many of these are my students in the media manipulation disinformation as well as my fellows and colleagues in at the shorenstein center and so good luck you're gonna have to you know you're really gonna have to bring your a game here okay i didn't know if you knew the forum was an adversarial debate event um i can't wait yeah but super hacker man what's the credentials here if i say up up down down left right left right can you finish the phrase uh bab start yes okay so you've passed and uh well the game though what game that's contra right and that's uh sort of an early how many early cheat codes 30 lives come on come on right but um i open with a little bit of a joke about you know the way in which we encounter technology in our lived experience something as a secret something as a playful thing something as a treasure sometimes something as a an organizing tool a workplace i mean come on how much of our time is spent on spreadsheets and other forms of documentation on these computers but they also bring a lot of levity a technology can bring a lot of coordination cooperation it can bring a lot of people together in some very positive and pro-social ways and then in some of the ways that your career has really uh grown up around which is in very dangerous and nefarious and in sneaky ways um but i would love to open up by asking you a bit about you know with a little bit of a shine on the your past careers at microsoft and uh the first version of cisa what has technology meant to you as you've navigated different professional roles so the way that's a broad question um i look i think whether it's in the the jobs i've uh been able to do or just the way that we experience everyday life i mean look the point of technology is to make things easier it's to make things more efficient it's to give us better telemetry and understanding and derive better insights the problem though is that the the balance of benefit and the downside um it there are downsides and that allows bad actors who want to monetize the technology pieces gives them plenty of opportunity that's that's that is like the best case study for that right now is ransomware right there are vulnerabilities that bad guys have figured out how to monetize and there are no meaningful consequences on them so why don't we address the vulnerabilities and this is really where i've spent a lot of time is on the policy side but why don't we address the vulnerabilities then well the problem is that daniel measler talks about this with kind of the gain loss or the win-loss ledger of software and and technology products like the benefits of technology still far outweigh the downsides even with hacking even with this information the benefits we derive are still so far out in front you know effectively security is our technology is in software it's as secure as it needs to be and so for now we're kind of left with this broad attack surface uh that that is continuing to lead to societal harms not just on the security side but on the erosion of trust and confidence in democratic institutions good answer i'm just being fun we're just having fun but let's talk a little bit about the the way the internet has developed um over the last uh at least 20 odd years as as we've started to enter this moment of more than just content delivery but connecting people to people and you had talked about a bit a bit about attack surfaces and in our course we also talk a lot about the internet being a tool a tactic and a territory now one of the precepts of our lab is always that everything open will be exploited so if you think about the structure of the internet infrastructure communication infrastructure in particular where do you even start think like securitizing something like that i mean do you start at the devices do you start at the components you know well so i this is again i keep using ransomware as an example because i think it's something that most people can relate to given what happened this summer and with colonial pipeline um there is there are no sources do you want to explain colonial pipeline just in case yeah so this or what was it late may early june a ransomware actor locked up the um the i.t systems the business networks of colonial pipeline which feeds from the gulf of mexico all the way up to new york harbor and it serves aviation gas and refined fuel products for cars gasoline throughout the eastern seaboard and to the point where i think it's one of the largest pipelines if not the largest for those refined products in the united states uh locked them up they were offline for you know several days which created not just and here's here's i'm about to get to the punchline um it's not just about the direct effect of an attack it's not just that the things were kicked offline and there was a disruption to the to the flow of fuel there was the attendant psychosocial impacts of the attack where not just was fuel stopped but there was fear amongst the population in the affected area that there was going to be a longer term outage so then that created a run on available stocks which you can take gas out of the ground faster than you can put it back in and we see this every year in hurricane season as well so that is why and there's a that piece of psychology that perception hack is what we really had to optimize for in the preparation of the 2020 election it wasn't just about protecting the election systems technically it was also realizing that there would be people out there that would use the narrative that a system was hacked to try to drive fear and undermine confidence that we also had to account for and that's what led to some of the initiatives that we developed and that might have a lot to do even with the way people made assumptions about what really happened in the 2016 elections and you had all of this fear-mongering around the influence of russia and the attack surface that was social media at the time and it does strike me that there are some uh similarities between these examples that you're bringing together um because it's it's always with the wires there's nothing i can do but the uh you know the way in which the internet is designed there's always a lot of digital shenanigans right and so that notion of perception hacking which i've heard you talk about a bit before sometimes we may over explain something as a very impactful technology as if online advertising is as as useful and as amazing as it would have us believe is that part of the perception hacking like ver v like uh uh tactic and their strategy that we should also be worried about i think so i think technologies um certain technologies are too complicated in the networks and the systems and then the systems of systems are too complicated for the human brain to actually comprehend and make decisions and trade off decisions based off of so one of the things about again about elections is that it's not really clear frankly how technologies are used to help conduct elections and in fact they make elections more accurate and they make them more efficient so when you think about counting the humans when single humans counting ballots for instance have an error rate somewhere on the order of anywhere from like 10 to 20 machine counting assuming no adversarial malicious interference it's like below zero not below zero yeah below zero i'm spreading this information one one harvard university all the time um we're all but no i hear you is that machines uh machines create a lower error rate and humans are fallible and i would say there's no such discipline more aware of this than sociology where i come from because counting things is hard right and that's something to be uh thinking about and humans are really bad yeah repeated tasks yeah uh over and over and over with consistency with consistency well we get bored we start thinking about other things you know like vacations other places to be what we're gonna have for dinner right like that repetition is is something that machines do a lot better with right because they're not wondering what their friends are doing you know or posting on facebook or like yeah figuring out how they're going to check their instagram and not get caught right right um let me ask you then about the critical infrastructure question because you brought up colonial pipeline and um what i love about being at the harvard kennedy school is we love to discuss really boring things right we love it and things that like infrastructure are it seems my class is laughing because they know how much i just love infrastructure and when we talk about communication infrastructure you know i'm a big fan of saying that on one hand i think andrew breitbart was right that politics are downstream from culture but culture is downstream from our communication infrastructure and for a long time though when we thought about critical infrastructure we're thinking maybe power grids and telephone uh maybe we're thinking of tubes and wires that bring us the internet but we're not necessarily thinking about the products online as critical infrastructure and how has what's been happening with social media particularly around perception hacking about what it means to vote in an election changed or augmented or should it augment our thinking about what should count as critical infrastructure so that was an arc of a question that i kind of want to go back to the beginning on um that i will ultimately answer your question but so i'll be here so there are critical infrastructures that uh so the u.s has you know it depends on the year or the month but anywhere from 16 to 17 to 18 different critical infrastructure sectors in the us we tend to over count or categorize you know our european counterparts only think six or seven sectors are truly critical we tend to think that anything that plugs into the economy supports national security or economic security is therefore critical uh we're getting better at really defining what the functions are rather than just the organizations like this bank or that bank but instead it's like it might not be the atm network of the bank but instead it's the wholesale payment system so really trying to figure out those things that are systemically important and that's that's where we have to go to really truly understand risk but there are things like power generation transmission and distribution that in any significant weather event hurricanes or otherwise are top of the list they're also very tightly linked with communications infrastructure so a lot of these things kind of run in the same pathways or use the same easements and so when you talk about restoration those things from a physical perspective have to be restored in very tight synchronization water is another one now the interesting thing so i'm going to use hurricane maria as an example here but when hurricane maria came through and wiped out the power grid wiped out the port or at least shut down the port one of the things that we really focused on restoring as quickly as possible was the communications infrastructure because of the psychosocial impacts right so that you could make sure that the people that were stranded on that island experiencing the challenges that they were could at least phone a relative phone a friend um but be in touch and say look i'm okay how are things up there uh you know we don't have power yet but at least i have a reconnection and even though we're on an island you know at least we have some better connection back to uh back to the mainland or elsewhere and so we actually worked with att the a few of the other um telecommunications providers to actually get sell on light trucks and sell on wheels so it just looks like a big uh truck with a big you know right dish on the back tower on the back and we put him on a c5 airlifted amount of dobbins air force base in atlanta and got him down there and that was just one example of you know you when you think about critical infrastructure think about the things that connect people um and then then you you at the same time have to worry about power that power takes a little bit longer so the definition of critical infrastructure is evolving over time i think the technologies that we use yes there is an i.t sector there's a communication sector i think the the fiber the you know the tubes is you said it connects us all that stuff is absolutely critical infrastructure but we also have to think about the systems and services and that are effectively now utilities like cloud infrastructure you know we have to start thinking about those things as utilities because of the aggregated risk that they present in the data centers and that that that mesh of regions presents to the way that businesses governments and others conduct operations so let me just dial in a little bit then because when facebook and whatsapp and instagram went down people were calling the police saying like facebook's down right so this isn't but is that is that a is that because they view it as critical infrastructure but you know like where where do we go with something like that so how many people here use a like facebook or something like that as an authentication tool right to get into other websites so if you want to log in to you know whatever you know the boston herald of washington post or new york times you can use facebook as an authentication mechanism like that okay and then you talk about now the passion's coming oh yeah come on triggered me yeah so instagram it's a massive massive uh commerce tool that's where a lot of small businesses market and how they conduct daily business operations like it was a significant outage for a lot of infrastructure and that's what we we don't always see those connective tissues that you know we just think about you know goofy dog pics and and you know beach pictures but there's a lot that rides behind that and i think that's where we need to have a national policy conversation on what those things where we're seeing concentrated risk that perhaps we haven't built in the necessary resilience uh and so it wouldn't take much it and we don't even have to be thinking about you know determined foreign malicious adversaries i mean the the the number one culprit of of uh power outages domestically is there's it's squirrels don't claim that don't blame them don't bring them into this the amount of economic loss caused by squirrels on an annual basis would shock you sure my my local telephone company is happy to hear it yeah yeah well let's double down on on resilience for a little bit because one of the funny things about um you know you have a very uh high-tech sounding you know uh job when you were at cisa and your director director but you know like do you say security twice cyber security infrastructure security agency we like security so much it's in our name twice but you know it's it sounds good it sounds a little technical but i only say that to bring up the fact that you know when we talk about resiliency and one of the jobs of cisa has has been under you especially was to secure this election but you're not out here asking people to vote on their phone right you're a big fan of paper ballots and is there um bruce here no no bruce come on we need triggers we had to walk out a while ago he was really very very triggered by some of this so you know but that's that's something where i you would think that you would in an ideal system technological system you would be able to secure it in such a way that people could be counted having one vote you would think that the system was that tight and like security ready but i don't think we're ready for something like that even though we're out here using our names and our credit cards and throwing our money around the internet but it doesn't seem to matter as much but there's something sacred about why so like tease that out right like what's the difference between your bank account and your vote right there's immutability with your bank account yeah i can call up who you are and how much money you have and what transactions you're conducting yeah if fraud happens though i've got a uh so so that's the difference right so with banks you you have that connection between identity and the the thing that's associated with it with with voting the as long as the secret a secret ballot the secret ballot is the law of the land in all 50 states you have to have the privacy separation here and you just can't achieve that the technologies aren't there uh right and if somebody says with the blockchain i swear um it's just it's it's it's not there so in this in the meantime you know we don't always need to be racing headlong into the next technological innovation because it makes things easier or more accessible we it's okay to dial it back and keep it simple all right so with with voting we want and this is so this is one of the biggest differences i think between 2016 and 2018 was we went from below not zero below 80 percent of votes cast in the u.s had a paper ballot associated

with an auditable record for the 2020 election it was 95 okay so we were we were able to work with states to either retire systems they got funding from congress and they retired systems they were touch screen with the votes saved down on removable media and there was no paper record associated with the individual votes and given that there's no way to meaningfully audit the vote count so whether it was from covet in states like new jersey that switched over to mail-in voting entirely from those old systems or states like pennsylvania and georgia that went away from those old systems to voting systems that use that mark up a ballot that gave us a lot of confidence in the ability to go back and count the votes so when the claims were in the 2020 election that the kraken or hammer and scorecard or whatever the communist algorithm it was it was dead venezuelan dictators that were adjusting the vote counts we were like okay well how about we count it so georgia counted it they counted their vote three times and it was consistent every single time and that's again paper gives you the ability to audit so where do we need to go from here we need as close to 100 paper as possible and then we need meaningful post-election pre-certification audits in all 50 states and we're about i think it's about 32 or some odd now not not post-election audits that happen you know six months after the fact you mean by my shadow cyber ninja security agency so there are standards for elect for audits for elections no there are best practices no people that do this you can't just uh throw them all out on the floor and say take a picture of this put it on social media say the whole thing's been rigged you can't just do that that is not a best practice it is in my book if i was running elections but again the point was the the point was resilience right that we could take a hit and as long as we had the fallback on an analog solution that we had comments so if georgia and pennsylvania had not made that switch to those other to the the new devices i would not have been as confident yeah in in 2020. well let's talk about taking the hit because you uh cisa was a little-known agency we had that we had people had been talking about it here and there and the lead up to the election uh i had been on some conference calls with other researchers people from sisu you were taking it pretty seriously you created a a website uh called rumor control you were gonna you know put little stubs up saying this is something that's been dispelled and then like you come out of the box a few day a few weeks after the election and you don't tell us that this whole thing's been rigged in fact you say the opposite right and you really took the fall um for democracy in my mind or at least the integrity of that election what was that like for you i mean it is in some regard easier to be quiet right but what did you feel like was there a duty you felt for your team and all the work you had put in uh did you feel like getting inside so you no yes that was in my bucket list getting fired by tweet well you amarosa i mean i thought i was watching the last season of the apprentice there so or at least the season finale the season for now um so look so you said you know we took it pretty seriously like i mean it's defending democracy like what else i mean can you take anything more seriously and that was the mentality across the team that if we do anything meaningful in the entirety of our careers or our lifetimes this is it like this is the thing and so we spent three and a half years working threat modeling trying to figure out every possible disruption that could be launched against the election and and it we used a foreign threat actor you know typically considered russia but you know they put the playbook out there in 2016 and so we knew that that others would be picking it up and running with it and you know whether it's iran or others um but the funny thing is that as we got closer and closer to the election as and this was kind of over the summer we we were thinking through you know what it's not the technical attacks that are keeping us up at night because we think we've got a a pretty resilient system with good indicators good networks that we'd spot stuff coming and we'd see it on maybe not on network but we'd see it through the election process but but this stuff is still almost black boxy and it wouldn't take a whole lot for someone to come out of the woodwork and you know because look disruptions happen every election day i think back to um the various um uh the primaries like i think georgia and virginia both had road work where backhoes severed fiber that knocked off a precinct and like stuff like that just like we've seen conspiracy theorists take and spin any other little minor story like that you know elections in particular yeah i mean it just really it kind of it ramps up so so it came down to the perception hack that we were most worried about again planning against uh a foreign adversary but when it came down to it in the middle of an election when it's all domestic based the the oath that we pledge to uphold and defend the constitution is is foreign or domestic and so we were within our authorities it was within you know the the oath and the pledge we'd made and it was not terribly difficult to to to call out nonsense where we saw it and it wasn't you know targeting individuals it wasn't targeting you know specific claims from individuals it was seeing themes emerge and watching the engagement watching the activity and providing not rebutting them on their face because that's just dumb it helps amplify the original claim but it's putting authoritative information out there on no a sharpie does not bleed through a ballot and cause either a missed vote or a vote for the opponent and i think that's really important about the stuff that we study when we talk about media manipulation disinformation campaigns you know there is quite a bit of calling out nonsense but there's also a quite a bit of it that is either state-sponsored propaganda when we look at these issues uh across different nations there's different things that get different wedge issues that get exploited whether it's religion or race or gender whatever is of highest contestation and polarization in different countries we will see different actors step into the fray whether it's government themselves or foreign adversaries or political pundits or just the entire grifter sphere what do you think could or should be done about situations like that where social media does play some role because of the amplification effects the as we've talked about in the past the velocity here but what else should we be thinking about uh you know feel free to bring up any of the things that have come up in the aspen information disorder commission you know you've got a very young group of people here many of which will go into uh policy work either public policy or public service or will even some of them will go into some of these companies right what do they need to be thinking about now so i think you you actually may have given up an easier case study of saying a foreign adversary getting on a platform and amplifying or manipulated data the way i have have viewed for some time now disinformation it's a supply and demand problem and on the supply side you have a disruption factor that the intelligence community can can provide by digging into their networks and their intelligence collection understand okay so this country is trying to do this and they're targeting this issue and they're using these identities uh in these persona all right let's roll that package up and then share it with uh a platform that's being manipulated and let them conduct their operation and their own investigation and then so those sorts of things and i think you know this is where you know to their credit a number of the platforms actually have done a a pretty good job on the targeted campaigns by foreign actors but it's also kind of low-hanging fruit and that's like a tiny little slice of the pizza you know well is it do we know what i mean we don't have the data oh so perhaps there's some recommendations okay that where you know we hear a lot of talk about regulating the the platforms and i honestly don't think we know enough about how the platforms operate right now to make meaningful regulation meaningful legislation to then inform regulation there's we have to have more required disclosures from the platforms the thing that i kind of compare us to is uh i guess i'm going to date myself here but it's actually did you have a live journal is that what you're going to about no no so when i was on myspace no i'm kidding um but back no back in back in the early 2000s with uh with enron right enron was a significant fraud and accounting uh scandal because well among other reasons there were they there was an appropriate transparency and disclosures on the part of the company that led to uh you know the collapse of this company that then also led to sarbanes-oxley which is a required set of disclosures for publicly traded companies so i think from a from a informa from a platform perspective social media platform perspective i liken us to a post-enron moment where there that we haven't had the appropriate disclosures and transparency into the business operations the financial models the the algorithms the moderation the targeting the advertisements the sponsored content all these things and so we need a sarbanes-oxley equivalent not to regulate content not to regulate speech but to demand a set of disclosures and access by researchers into how these platforms conduct their operations and that will do two things i think one is provide more insight into for instance what are the desired you know what does success look like for how a algorithm is optimized you know what are the sorts of outcomes you're trying to achieve and that i think would provide the market a little bit more understanding of how they're being gained and they can make a set of decisions so recently certain disclosures made by reuters over how um at t and directv sponsored one america news like that was like oh well we always knew they were there i personally don't like how they've amplified um election disinformation and so i had more information where i could go make an informed decision with my dollar so i switched off att and i canceled directv good so that you know there is more information transferring more time for youtube yeah i i signed up for the youtube live tv um but so so one is like it gives the market more information but the second is is it gives policymakers more information to make the appropriate policy interventions we don't i don't think that we have enough information right now to make meaningful targeted policy interventions yeah and this goes back to needing many much more research on the harms caused by these platforms including just including just basic like financial fraud personal injury and then of course there are these collective social injuries like on january 6th but we're about to go to audience q a so i just want to alert you that people can line up keeping uh a healthy distance between one another at microphone there and over there and then up there and over there and we will take questions from the crowd but earlier i mentioned pizza do you like pineapple on your pizza i do not no well what's the deal why did sissa get involved in that war because that you know that's gonna that's a big one so 2000 summer 2019 we launched a campaign called named the war on pineapple so the idea here was remember that supply and demand problem of disinformation so there there are parts of the federal government that have the opportunity to engage on the supply side and disrupt supply civilian agency like cisa which is a public-private partnership and education outreach and engagement oriented agency has to focus on the demand side and so what we were aspiring to do was help increase awareness on how disinformation campaigns influence operations work and so we distilled it down again this was in the wake of 2016 and some of the techniques that the russians used to amplify kind of social discord but we distilled it down to five steps the first is just identifying the issue the second is seasoning and putting accounts into place to start amplifying the issue third is actually amplifying it fourth is taking it mainstream and so that means getting it out of the you know the the 4chan 8 con whatever and and then even off of facebook into the media where you actually want newsrooms talking about the thing you're stoking and then lastly you want to bring it into the real world you want real live in meat space people protesting and counter protesting and they were the russians were able to uh get count protests and counter protest on race issues on gun guns rights issues and that you know so so we used we had to find an issue it can't be you know political issue because you turn off 50 of the people in a in a bad way like lizard brain shut off type um and and then you know same thing with like you can't talk about russia because then it's like russia russia russia uh so the team uh the election integrity team uh the election security initiative team went out to lunch and they were debating like okay what are just like normal type things that are kind of binary where you know it's like you're you're on one side or the other like cilantro cilantro or salt and vinegar chips or black licorice um now twizzlers are awesome but not black licorice but no it came like the most animated discussion at the lunch table was hawaiian pizza whether you like pineapple on your pizza or not it was like if that like a star was born like in that we launched this campaign i was in new york went to a pizza uh joint and had you know a slice of pepperoni and a slice of uh hawaiian and we we launched it we were working with um secretaries of state and election directors and had them pitted against each other so it actually was honestly a coordinated and authentic behavior campaign so we probably should have been yes we probably should have been moderating no you just you labeled it you put your name on it it was we can also have you know public public uh what are they called psas public service announcements but it took off and like now anytime hawaiian pizza comes up on twitter i get tagged and it's just it never ends and you really don't like pineapple on your pizza it's just like but like pineapple cheese and tomato sauce like together oh well you haven't lived apparently we have some open microphones so please who's got courage tonight one of you does but we've done a couple other while people move to the microphone we've done other sort of kind of innovative different sort of um activities and campaigns to generate awareness we there are a couple graphic novels uh on disinformation the first one was real fake and the second one is bug bite bug bites those are both available on sisa.gov and then the third was a was a game called harmony square where you could actually play both sides but you know you could you know the objective is to actually cause chaos in a community through disinfo okay we've got a few intrepid people so name affiliation at the school here at harvard and then how much i paid you to be a here that ends in a question mark um hi um my name is andrew i'm a student at uh professors uh professor donovan's class i'm a master of public administration student here um professor this you you're hilarious this is your best life yes um you should do all of them uh chris question um so you talked about pay him well i do get them all pineapple pizza at the end of the quarter i was promised five um chris you spoke about disclosure and i was wondering uh by the tech companies i was wondering what the status is for this i would assume there's a lot of resistance because algorithms stuff like this is intellectual property it's part of their dna they don't want to let it go um what the status quo is for disclosure in terms of the state and what kind of levers you have for coercion or persuasion thank you it's gonna i mean i i don't think self-regulation works there have been plenty of opportunities for self-regulation to date and there haven't been meaningful cost impositions to change behaviors i don't think so it's going to require legislative action um which you know if you're as jaded as i am you you think that that's really difficult in today's political climate in congress but if there's anything that both sides of the aisle hate right now and it's it's social media platforms and they're willing you know they to you know they'll hold their nose and take lobbying you know take political contributions but i think that this is an area that that you can probably find some middle ground it won't be a dramatic legislative package but there'll be some kind of bare minimums and some of the proposals i've seen you know bare minimum access to um certain you know whether it's content moderation libraries you know the again the the kind of the parameters for uh algorithmic optimization you know opening up opening that up to a limited set of security researchers and journalists um to to review i mean again on their own they're not going to do it so that's that's just but that's just table stakes for me there's got to be i think a lot more so i think something will get done it'll disappoint everybody but it'll move continue to move the ball forward and i don't think we should discount the possibility of culture shift right it's not that you had to legislate friendster and myspace to go away sometimes something else just comes along so that's a betamax right the vhs like technology always has these moments of replacement this so the betamax vcr thing that's an entirely different jfk junior forum um and why vcr which is the inferior technology um prevailed uh so we had both i had a vhs handlebar um so we have to you know to your bigger point like the the social media platforms we have today may not may not and will not be the social media platforms we have tomorrow and so i think to the market driving i mean i do think there is an opportunity for differentiation um and that somebody actually making a uh socially responsible platform going forward now at the same time we're going to see even crazier vulcanization and fragmentation of the social media market well you'll get you'll just see crazy stuff over here but you know one thing that i think is also open for government intervention is it's i think it's about time for a digital services agency or a digital agency they can think about privacy focus on privacy they can focus on trust and safety programs because we still see a very you know inconsistent landscape for trust and safety measures and platforms um this is where we get into like won't anybody think of the children though right i mean come on we just gotta have apps that children can't use don't you think like resorts that children can't go to so you can have a nice vacation spoken like a lesbian i know i know but i i do agree with you so there's but there's again these there are a lot of table stakes out there i just think that from a bureaucratic perspective we still there's a lot of internal resistance and a lot of kind of money on the sidelines that are going to keep things in check or keep things in place go here and then we'll go over here hi thank you so much for coming to speak with us my name is rachel i'm a junior at the college i study folklore mythology and my question is um you've worked on top-down efforts to monitor and combat misinformation how effective do you think bottom-up activist community organizing efforts are and how what would that ideally look like so what are the problems right now with um particularly way the some of the 2020 election oh gosh here's the chance yeah um one of the problems with the the 2020 election disinformation is that where it was initially top down it actually and kate starboard talks about this out at uw but in others but participatory disinformation where you actually see this rotation back and forth and i think what's happened is that it's metastasized to a level that there's no longer control it's actually it's it's gone i think to really kind of to your point it's gone almost to an astroturf operation where things are starting and you've lost the ability to control what you thought was a campaign that was going to serve your own own benefits and that's and unfortunately that's that is my biggest concern about the run-up to 2022 in 2024 beyond that is that we've lost the narrative and that has always been the point by the way that is always that was always the point of the russians in 2016. that

was always the point in 2020 it was to have the rational thinkers lose the ability to understand what was what the truth was what the facts were to doubt the system and cause chaos and therefore you don't know what's true anymore yeah and so that's that's where we are and you know when you have you know virtually an entire political party embracing that trend as i've said on face the nation so i'll say it again here it's it's an anti-democratic death spiral yeah and to that cutesy metaphor um of the anti-democratic death spiral you know the the way in which grassroots organizations are shuring up their defenses jonathan corpus song is here somewhere where is he he's over here so he's been researching this in the context of the philippines where uh you know civil society organizations actually have to dedicate a good share of their communication resources to fighting it out on social media and to into calling out these campaigns and saying these horrible things are happening within the us there's a coalition of about 200 organizations called the disinformation defense league that pool ideas and resources and sometimes work together on social media campaigns but it's really just um you know the slingshot versus goliath at this point right because of the resources as you're saying that go into the coordinated campaigns where political pundits or politicians are willing to burn their accounts as well especially if it means gaining notoriety on particular issues good for fundraising it is incredible for fundraising and just to think that all these school board protests are organic like that's the point of astroturfing no these are being coordinated and sponsored incredible over here great question thank you how's it going everyone i'm derek i'm an mpp2 here at hks uh so i have a question kind of about your point um about how little we know about digital platforms i think i agree that you know it's difficult to make meaningful um and good legislation or regulation around them without knowing more but to the point about sort of pushing for transparency and then using um knowledge and information to produce based on their disclosures to make laws at some point in the future like producing knowledge can take a long time it would take you know years for studies to be done for a sufficient body of knowledge to be built to make informed decisions and then when that happens there would be a lengthy public debate process and in between now and then there would be elections and coups and genocides um all things there will be like facilitated or impacted by disinformation so like in the immediate term what do you think uh sort of we as policymakers as citizens um as a society can do to sort of uh guard against disinformation i i mean this is this is the question right there's no silver bullet there's no single solution we have uh you know we still have freedom of speech issues and the government you know the last possible thing i would want is the government start wading into the middle of content moderation decisions and start declaring things off you know out of bounds or off limits and we really struggle with that i think in both the research community as well as you know in the policy maker side and that's what's held back that's what like when it's the russians interfering with elections oh that's easy oh you can deal with that but when it comes back home and and particularly when you talk about a lot of the domestic extremism issues but look free speech is not absolute there are you know there carve outs for defamation there you know there's also criminal issues here so uh i i think given what's in front of us particularly with the facebook whistleblower disclosures we have an incredible resource that i think one of the first imperatives is sorting through that find out what we have and that can help jumpstart some of the you know as we wait a year and a half for legislation and then regulations after that use what's in front of us to help jump start maybe see around corners a little bit faster we can hope so and you know i think that there's you know quite a bit of journalists and other folks working on this right now who are trying to get to the the nut of what facebook knew and when they knew it especially around violent and citing content so one more piece just because that's like a completely unsatisfying answer and i get that uh because you know what i am fearful of having been in government now a couple times and seeing how they you know the government's both ineffective and bureaucratic and actually oftentimes makes the wrong decisions the other thing i didn't really emphasize here is is you can't like don't for the love of god do not rely on the government to fix everything there are still things there are there is there are table stakes that we can do so professional organizations medical boards state bar associations have to self-police they have to take care of themselves and their own members and and call out and kick out people that are abusing the the you know the trust and abusing the privileges that come with membership we saw that in oregon recently where uh there was some some coveted related disinfo um and then you know we continue to see a series of bar disciplinary actions across the country because of election disinfo claims over here i am a senior research internet policy fellow at jones uh with jones team at the shorenstein center my name's april so there is a real urgency to do something now and a lot of verve in a lot of different directions to get something done but what are the dangers of doing something wrong now or getting the policy incorrect and possibly codifying something that will be really hard to undo when i ask that in light of the fact that i see platforms saying they want to be regulated right which is always kind of a red flag because the things that they're calling for wouldn't necessarily curtail their power and so i'm curious like you know the midterms are basically tomorrow people want to do something now makes sense what are the dangers of getting it wrong so i think without developing you know having an open policy discussion and debate you know you run into a situation where with significant unintended consequences that have civil liberties and free speech impacts um that is a number of the recommendations i've seen you know we start losing um some of those freedom of speech protections and and that you know i think in the meantime there are things we can do in terms of limitations like section 230 you know it's not like throw section 230 out which gives immunity to uh anyone for for the platforms for for content moderation purposes but there should be caveats there's there should be there can be carve-outs particularly when you talk about financial in advertisements and once there's a profit motive associated with um you know and the ftc for instance has authorities right now for fraud and fraud enforcement so you think about uh dr mercola and some of the others that the the pandemic profiteers we need to resource those that already have the authorities given the capabilities to actually more meaningfully uh enforce against it that's that's where we start yeah enforcing even the uh what they have already on the books in terms of terms of service what was horrible about some of these disclosures is you figure out that there's millions of people that have very high influencer accounts um on white lists that basically don't even hold them to account for the um terms of service right which then provide cover so that's where it should be the other way around yeah if you have more followers you should have more responsibility right yep more amplification more more responsibility you know that engagement drives the business model and drives more clicks and yeah that's ad revenue yeah and you know and section 230 i've always viewed as a a policy of d control that is it was put in place at a time when you didn't have instantaneous high fidelity broadcast and download but now each and every one of us is like a television station you know with our telephone and the advertising model wasn't created yet and the advertising model wasn't baked in directly which allowed you to to monetize everything so instantaneously and i think it would be a good bet to push uh for regulation that would at least regulate political ads in such a way so that there was more transparency and we knew how much was being spent and do mandatory reporting rather than these volunteer you know transparency databases i don't disagree with you it's never going to happen yeah you don't disagree no because the ones that you're actually going to impact through that regulation are the political the people that you're asking to govern themselves right and that just doesn't work that way that is that's true well unfortunately we have reached the end of our sparring session so i think we get to bare knuckle box in the alley oh is that right yes i think so are you up for it sure we didn't disagree really on anything it was a great conversation chris was fun it was really nice to talk to you thanks for watching and uh i'll remember next time not to put pineapple on your pizza and please and thank you for what you were willing to do it took a lot of courage to stand up in that moment and to fight for your team and to fight for the work that everybody had put into that election and we are very thankful and grateful to you for being willing to to really take that hit so thank you so much thank you you

2021-10-23

Show video