When "Tech for Good" Goes Bad
Lessons from the failure of Raheem AI, whose founder sought to abolish policing, but got caught using his nonprofit's funds on personal luxury. Plus two things you can do about the election, now.
Two quick notes on what you can do if you’re anxious about the election (who isn’t) and then a longish post about ethics in civic tech.
1. In the nine battleground states that will decide the winner of the Electoral Vote, local organizing groups are still in need of funding. The Movement Voter Project, which has a great deal of experience raising and channeling funds to these groups, estimates that they have a total funding gap of $200 million to $300 million. That may sound like a lot, but the Harris-Walz campaign raised $540 million in the single month after President Biden passed the torch.
When you give to MVP, the money goes to groups that do year-round organizing. Right now in many states, it is still possible to register to vote for November. Those groups shift to voter persuasion and mobilization efforts as voting begins, and then also do voter protection work up through and after Election Day. Also, by supporting local organizing groups, you help them build the power to convert election victories into policy shifts. One of the reasons this election is so close is the many voters who don’t see tangible results afterwards, so giving to year-round organizing rather than individual candidates helps chip away at that long-term problem.
For a deeper dive into the details, you can read this September 6 memo from MVP’s founder and executive director Billy Wimsatt. There’s a long list of local groups organized by state, if you want to focus in a particular place—or you can just donate via MVP and let them sort out the details. There’s also an online briefing tomorrow at 3pm ET; register here.
2. Want to help register more likely Democratic voters? Then check out Civitech’s New Voter Initiative, an interactive map of the USA that targets nearly five million people they’ve identified who aren’t in the voter files used by Democratic party vendors but who they’ve scored as likely Dems. For just $2 per voter, Civitech will mail a voter registration form and include a pre-paid return envelope where possible.
Civitech CEO Jeremy Smith tells me that this list “is sourced from more than ten thousand separate records locations and matched against typically weekly voterfile updates. We score these individuals in house and perform validation surveys by the millions, along with independent model audits by Data for Progress and others. So, these are carefully curated lists of likely Dems, not currently registered.”
In 2019-2020, Civitech did something similar, which generated about $11 million in contributions to various organizations that added the map to their existing programs, and resulted in more than $2 million being spent on voter registration mailings. Of that, Smith said more than 400,000 people registered, three-quarters of which then went on to vote. Post-election surveys and observations suggest 89% voted for Democrats. If your organization wants to add the map to its website, contact Tracy Davis at Civitech, their VP of marketing.
Good Intentions, Bad Actor?
A little more than two weeks ago, The New York Times published an astonishing story (gift link) by David Fahrenthold on Raheem AI, a recent civic tech start-up that sought to transform policing in America but appears to have mainly spent hundreds of thousands of dollars on its founder’s private needs. That included, in the year 2021 alone, $80,000 on luxury vacation rentals, $46,000 on ride-share services, and $11,000 on upscale clothing purchases, according to internal records. Even more astounding, Fahrenthold found that Brandon Anderson, the organization’s founder, had apparently invented the story of its namesake Raheem, who Anderson claimed was a Black teenager he fell in love with and was set to marry, but was killed in the early 2000s by Oklahoma City police after being accused, unfairly, of driving a stolen car. No record of any such person or killing could be found.
Anderson was “a handsome man with retro browline glasses and a touch of gray in his beard, [with] the look and optimistic tone of a Silicon Valley founder,” Fahrenthold wrote. Well, not quite, since he was Black and Silicon Valley VCs barely ever bet on founders of color. But Anderson was never after VC money; the return on investment he was pitching, mainly to cause-related accelerators and foundations, was fundamental social change. Even more, the story he told about the love of his life and how his supposed killing changed him and gave his life a mission, clearly sold many funders on his nonprofit, which raised more than $4.4 million from places like the Kresge Foundation, FastForward, Echoing Green, the Skoll Foundation and the Dropbox Foundation.
At first, Anderson set out to build a website that would enable people to file complaints against the police from their phone. His boyfriend, he claimed, had been killed by an officer who had long been abusive but who had never been subject of an official complaint. If more people filed such complaints (with the help of free lawyers his start-up would provide), then Raheem could use their stories to lobby for policies that would defund police and invest in communities. The “AI” in Raheem was supposed to be a chatbot that would assist people with filling out their complaints.
With this pitch, Anderson was embraced by some Oakland-based anti-police activists, profiled multiple times glowingly in the business/tech media (“the Waze of reporting police,” Inc. Magazine called it), incubated by the FastForward nonprofit tech accelerator (which threw in more than $700,000 between 2017 and 2020), made a 2018 Fellow by Echoing Green, another influential program for nonprofit innovators, and given a plum slot at TED 2019 as a Fellow. Then, in 2020, after the murder of George Floyd by a Minnesota cop set off a wave of protests and demands for fundamental police reform, Raheem got some rocket fuel in the form of $1.6 million from foundations suddenly searching for ways to show their support for the issue. By then, as Fahrenthold notes, Raheem’s original notion of becoming “Yelp for Police,” wasn’t working. Its website had collected more than 2,700 stories from users, but Anderson and his team couldn’t figure out how to offer a one-stop way to file those complaints with the relevant police departments.
It’s at this time, Fahrenthold writes, that Anderson started spending heavily on his own personal needs, but it’s not totally clear from the story if that was the case—or if records earlier than 2021 couldn’t be found. In any event, while apparently enjoying a lot of luxurious downtime and shopping trips, Anderson decided to “pivot” – another move loved by Silicon Valley evangelists – and announced that now Raheem would instead focus on building an app that would reduce the need for police entirely. How? By building a network of alternative responders and “liberated dispatchers” that it called the “PATCH Network” and then connecting callers in need with the appropriate non-violent community resources instead. As best as I can tell, some local alternative responder groups did join this network, but I had no success getting any information from them about whether the new version of Raheem made any difference in their work.
Nevertheless, the Kresge Foundation, which had given the first version of Raheem $220,000, now pledged $675,000 to Anderson’s new vision, listing the group among 59 selected for its “racial justice commitments.” “These are some of the most skilled organizers and racial justice experts that I have ever met,” said Trista Harris, president of FutureGood, a consulting firm that engaged many of Kresge’s grantees in its the Racial Justice United Network. Fahrenthold asked Kresge why it funded a new group with no track record to tackle such a complex task? Its spokeswoman responded that support to Raheem “did not come with expectations of meeting project deliverables.”
How do you get some of that?
Another group that backed Raheem through this period was New Media Ventures, which I’ve written about multiple times here at The Connector. NMV has been around for more than a decade, and by 2020 was fielding pitches from hundreds of start-ups hoping to get its support. To get backing from NMV was then and remains now a big deal for tech-for-good start-ups, as only ten to fifteen make it through its gauntlet each year. In 2020, NMV had 1400 applicants (a larger number than usual because it had expanded its open call in the wake of COVID); interviewed 100, and invested $1.2 million in just 13, including Raheem. Though the amount wasn’t huge—somewhere between $50,000 and $100,000, I’m told—a NMV endorsement also comes with additional coaching and networking support. It is meant to open doors to lots of other individual investors. In November 2022, NMV featured Anderson at its annual summit, giving Anderson a keynote speaking slot.
From what I understand, Anderson had actually pitched NMV two previous times before it backed him in 2020. The reason he didn’t get support earlier is because his initial idea for Raheem didn’t make enough sense—there wasn’t any way that enough people would pre-download an app to report negative encounters with police for it to scale to a meaningful level. Nor was it clear how the organization would get those complaints back to police. “It was kind of a two-sided marketplace that did not make sense at all,” one insider told me.
On the other hand, NMV was charmed by Anderson. He seemed like a “compelling,” “driven” “fail-fast kind of guy” who was trying to build interesting tech and who cared about justice, this insider said. The sort of person NMV was looking to bet on. So in 2020 when he came back to them with a fresh idea that, even better, was oriented toward police abolition rather than reform, NMV was inclined to give him a shot—even if his new idea might be even harder to pull off at scale than the original one.
As for due diligence, my understanding is that NMV’s approach involves tapping its own network of advisors and contacts to get a read on potential investees. It may also involve reviewing internal financial statements, but if an organization is fairly new it won’t have audited financials—something that might have caught Anderson’s hands in the cookie jar. It would never have occurred to NMV to question whether Anderson’s whole personal story about a former boyfriend killed by police had ever happened.
Until now, NMV hasn’t said anything formally about its role in boosting Raheem and Anderson. After the Times story came out, I asked Phillip Sanders, a partner at NMV who helps lead its investment team, what due diligence it put its potential investees through, and if it had any concerns about Raheem’s claims about its services, or about Anderson himself. Sanders replied with the following statement on behalf of the whole NMV team:
“You're correct that New Media Ventures initially gave Community Response Works (CRW), formerly Raheem AI, a grant as part of the 2020 Expanded Open Call. According to the diligence notes we reviewed from the NMV team at the time, in 2020, they were still a platform for reporting and rating police encounters and had expanded their theory of change to include policy advocacy via police oversight boards. From the notes, the NMV team was impressed by their team’s commitment to learning and adapting their approach at a time when community activists, political figures, and society at large were demanding tangible change in the wake of George Floyd’s murder.
In early 2022, while a part of the NMV portfolio, NMV gave CRW a small follow-on grant as they further built out their PATCH network of community health care and social service providers and abolitionist organizations, which served as the foundation for the PATCH technology (what you probably know as the alternative network of responders to 911).
NMV has always firmly believed, both then and now, that policing in the US is in desperate need of disruption, with potential rewards that greatly outweigh the relatively small amounts currently being invested. Data collected by Mapping Police Violence shows that more people were killed by police in 2023 than in any other year in the past decade–1,247 people. What cannot be lost in this debacle is that funders were and are right to pursue a solution to this specific problem and that a failure should not dissuade or negatively impact others who are working in earnest to re-imagine public safety in the US.
We’ve always believed in supporting bold, audacious ideas that tackle systemic issues, and as early stage investors and funders in the progressive space, we have a mandate to do so. CRW’s vision of leveraging tech, advocacy and organizing to build infrastructure and increase safety for Black & Brown communities resonated deeply with our mission. The founder, Brandon Anderson, was a compelling figure, someone who clearly understood the problems being addressed and who was passionate about creating a more just and equitable system.
To the question about our due diligence, our initial due diligence process is designed to be rigorous yet trust-based, with detailed reviews of an organization's team, financials, mission alignment, theory of change, and conversations with key stakeholders and experts. At the time of our initial grant to CRW in 2020, the team did not identify any immediate red flags during their evaluation of the nonprofit. And in early 2022, while already in our portfolio, their operations resembled similar tech-enabled organizations going through a pivot while heavily relying on philanthropic funding. However, in light of recent allegations, our current team is re-examining the situation to understand any gaps that may have been overlooked.
We understand that this incident may cause some to become more cautious in their philanthropic efforts, which is a response we hope funders will mindfully reject. Funding in this space is still crucial. The underlying problem still persists; there has been no real scalable solution that dramatically reduces police brutality in Black and Brown communities. The challenges we face as a society are too significant, and the need for transformative solutions is too urgent. While we must be vigilant in our due diligence, we must also continue to support bold, innovative ideas that have the potential to create lasting, systemic change.”
This all seems fair enough. Humans make mistakes, and the best of intentions do sometimes go awry. This certainly isn’t the first time public-minded funders or philanthropists have made a bad decision, either. And given how so much money already flows toward projects that simply reproduce the status quo, the last thing anyone wants is for the whole field to become more risk-averse. But Anderson’s success at fooling many, many people over several years ought to still give reason for pause. He convinced a lot of would-be gatekeepers to give him a glowing write-up or recommendation and parleyed all of that into real money.
Anderson wouldn’t answer Fahrenthold’s detailed questions, only providing a written statement that some of the allegations against him were “rife with untruths.” He also appeared to deflect attention away from his personal expenditures towards Raheem’s larger failure to achieve its mission. “It’s easy to assign failure to one cause or another in hindsight, and individual expenditures are easy to mischaracterize without the burden of context,” he said in the statement. “The bottom line is simply that it didn’t work, and as the leader of that effort I share most of the blame.”
I have a slightly different theory of this case. Anderson’s failure is almost a predictable by-product of the whole “for good” ecosystem. The various organizations that lifted Anderson up along the way—Fast Forward, Echoing Green, TED, the Kresge Foundation and New Media Ventures—all need people like him to justify their own existence. “Discovering Tomorrow’s Leaders, Today” is Echoing Green’s promise. “Through our full-time Fellowship, we find people working on plans to make the world better in a big way. Then we help them become impactful leaders by connecting them with the tools, resources, and communities they need to bring their ideas to life.” Fast Forward exists to “accelerate good.” “As the first and only organization focused exclusively on scaling startups that combine the best tech with sustainable nonprofit business models, Fast Forward’s mission is to accelerate impact.” At TED2019, we […] search[ed] for deeper meaning, exploring technologies that evoke wonder, mind-bending science, awe-inspiring creativity and, most of all, the possibilities that happen when we ask what ideas are truly worth fighting for, worth living for.” Etc. I’ve written copy like this myself.
People who work at these organizations need to have successes to showcase. And winning grants or “investments” validate founders, too. I can kind of picture Anderson getting a lot of praise in the early years of building Raheem, and him enjoying all the plaudits. But at more than one point in his journey, it must have become clear to him that he couldn’t meet the promises he was making to these audiences. The stress and pressure of being an early-stage founder, even one with a working idea, is enormous. And then you start telling yourself that self-care is important, and since you raised the money, isn’t it okay to use some of it to sustain yourself?
Beyond the need for more stringent due diligence, maybe the lesson of Raheem’s failure is this: the whole “for good” sector and especially the organizations who claim to be key accelerators and facilitators ought to stop demanding miracles from start-up founders. Maybe the people who are more modest in their claims and goals are more deserving than the ones with the most audacious dreams. And finally, maybe there really isn’t “an app for” fundamental change.
Also of note
—This Friday, from 2:00-3:00pm EDT, New America is hosting an online conversation with Audrey Tang and Glen Weyl on “How Technology Can Reinvigorate Democracy.” Register here.
—The Trust for Civic Life is looking for nominations for its 2024 Civic Entrepreneur grants. These $25,000 grants are for individuals and small organizations leading experimental new programs that get their community actively participating in shaping a community vision, identifying challenges, and working together to solve them, with a strong focus on the Black Belt, Central Appalachia, Southern Border, and Tribal Lands. They are taking applications on a rolling basis but will give priority attention to those that arrive by September 25.
—I always want to read Rebecca Traister on practically any topic, but her new piece in New York magazine is a must. It’s about how women and especially Black women, rallied around Kamala Harris and lifted her past a host of Democratic party leaders who not only wanted Biden to step aside, they wanted to shunt Harris aside too. Traister does an excellent job of chronicling the movement moment that started the Sunday night of Biden’s announcement with more than 84,000 women on the Win With Black Women call, and shows how that energy grew and cemented Harris’ claim to the prize. Traister’s article comes with a warning: “As we settle into the second phase of this candidacy and old hands regain control in preparation for the presidential debate on September 10, the question is whether the cautious, moderating forces that have long guided Democratic electoral politics will tamp down the people’s power that was unleashed this summer and jeopardize Harris’s chances of victory.”
End Times
If you want to share how you’re feeling during tonight’s presidential debate, go to https://jotto.me/cloud/debate2024 – that page will create a real-time display of people’s texts, audio and video.
Word, recounting of the Anderson scandal was done, brilliantly, and with great care. You are much closer to the nonprofit world than I am, and I was fascinated by your dive into it, and I have certainly resonated for years with your theory about the for good funders, apps, foundations, etc. Maybe that explains why I have had so much trouble finding the killer app for revolutionary change. If you stubble upon it, will you send me the link? Anyway, because as usual top drawer stuff and really informative and not that surprised.
I was interested in funding the vote map and attempted several times, only to have the system reject my phone number over and over again. So, no funds from me... Oh well...