Cinder Technologies
Last reviewed
Jun 3, 2026
Sources
8 citations
Review status
Source-backed
Revision
v1 · 1,358 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 3, 2026
Sources
8 citations
Review status
Source-backed
Revision
v1 · 1,358 words
Add missing citations, update stale details, or suggest a clearer explanation.
Cinder Technologies is a New York based software company that builds trust and safety tooling for online platforms and, increasingly, for artificial intelligence products. The company, which usually goes by just Cinder, describes its main product as an operating system for trust and safety: a single platform where moderation, investigations, policy enforcement, and abuse monitoring all live together instead of being scattered across spreadsheets and ad hoc databases. Cinder was founded in January 2022 by a group of former Meta and Facebook staff who had spent years fighting hate groups, terrorist networks, and state backed disinformation campaigns, and several of them also came out of Palantir and the US government before that.[1][2][3]
Cinder drew fresh attention in May 2026 when it raised a $41 million Series B round led by Radical Ventures, bringing its total reported funding to roughly $55 million. The round landed at a moment when the trust and safety category, once treated as a back office cost center, had become central to how AI companies ship products safely.[4][5]
When Cinder launched publicly in December 2022, it pitched itself as the first integrated platform built specifically for trust and safety teams. The founders liked to compare it to what Salesforce did for sales or what Zendesk did for support: take a function that had always run on a mess of Google Docs, PDFs, static block lists, and one off SQL queries, and give it purpose built software.[2][3] CEO Glen Wise put the problem bluntly in early interviews, saying that platforms "would hire really smart people but they'd be stuck in spreadsheets" because no tool had ever been built for the work.[2]
The platform centralizes several jobs that trust and safety organizations have to do at once. It handles policy setting and codification, case and investigation management, content moderation and reviews, and risk monitoring and compliance reporting. The idea is that a moderator or investigator can pull together user generated content, account metadata, and off platform signals in one interface, make a decision, and have that decision feed dashboards and transparency metrics, rather than switching between half a dozen disconnected systems.[1][3] In a launch statement, Wise framed it as an "end-to-end platform for centralizing operational data, streamlining decision-making, measuring impact and promoting transparency."[3]
Over time the product leaned harder into AI. By 2026 Cinder was describing its job as defending platforms from AI powered abuse, and it had layered AI agents, human review, quality assurance tooling, and expert services into the product so customers could enforce safety policies at the speed and scale that automated attacks now demand.[4][5] That shift tracks the broader move in the field, where generative tools have made it cheap to mass produce scams, fake accounts, abusive images, and prompt based attacks, and where the companies building large language models need to red team and police their own outputs.
Cinder's founding team is the company's main selling point, and it is unusually deep for a startup. The four people most associated with the company are Glen Wise, Phil Brennan, Declan Cummings, and Brian Fishman. Wise serves as chief executive. He had worked as a red team engineer at Meta, and earlier spent time at Palantir and in the US government. Brennan, who took on the chief operating role, brought about a decade of US government experience along with his Meta work on threat intelligence. Cummings ran engineering and also came through the government, Palantir, and Meta pipeline.[1][2][6]
Brian Fishman is probably the best known name on the roster. He spent more than five years as Meta's director of counterterrorism and dangerous organizations, the team responsible for keeping groups like terrorist networks and organized hate movements off Facebook and Instagram, and before that he had led disaster relief work at Palantir.[6][2] At Cinder he handled business development in the early days. The shared thread across the team is that they built the threat intelligence machinery Facebook used against some of the largest abuse campaigns of the last decade, and they started Cinder to package that capability so that smaller teams without Meta's resources could do similar work.[6]
That pedigree also explains the company's posture during real world crises. In November 2023, during the Israel-Hamas war, CNBC reported that the ex-Meta founders were selling their trust and safety technology to platforms scrambling to handle a surge of violent and extremist content, exactly the kind of fast moving, high stakes scenario the team had handled inside Facebook.[7]
Cinder has been backed by Accel from the start. The firm led a roughly $4 million seed round in May 2022 and then a $10 million Series A in December 2022, with Y Combinator participating, for a combined $14 million when the company came out of stealth.[1][2][3] Cinder had gone through Y Combinator, which is how the YC connection started.
The Series B in May 2026 was the larger and more closely watched raise. Radical Ventures led the $41 million round, with participation from Accel, Y Combinator, Microsoft's venture arm M12, PSP Growth, and Neuberger Berman's Outpost Ventures.[4][5] AlleyWatch, TechCrunch, and other outlets reported the round in mid May, with the deal dated around May 13, 2026, and pegged Cinder's total reported equity funding at about $55 million.[4][5] Coverage of Radical Ventures framed the investment as part of a wider push by the firm into applied AI, alongside other large rounds it backed around the same time.[8]
By the time of the Series B, Cinder said its software helped protect more than three billion users across its customer base. The platform's named customers include AI companies and consumer apps alike: OpenAI, Spotify, ElevenLabs, Black Forest Labs, Midjourney, Character.AI, Stability AI, Synthesia, BeReal, Depop, and Zello, among others.[4][5] Ben Brooks, head of public policy at Black Forest Labs, said Cinder "provided rigorous adversarial testing that matched our release velocity" and helped the team find edge cases before launch, a quote the company has used to describe how it works with model developers.[4]
Cinder sits in a part of the industry that has grown quickly as generative AI has spread. Building a capable model is one problem; keeping it from being used to generate deepfakes, abuse material, fraud, or harassment is another, and it is one that regulators and the public increasingly expect AI companies to solve. The wider field of AI safety tends to focus on aligning models and preventing catastrophic misuse, while the trust and safety layer that Cinder works in is more operational. It is about the day to day grind of detecting bad behavior, investigating accounts, enforcing rules, and documenting the whole thing for compliance.
A few things make Cinder notable in that space. Its founders bring direct experience from one of the largest trust and safety operations ever built, at Facebook, which gives the company credibility with buyers who know how hard the work is. Its customer list pulls in many of the better known generative AI startups, which positions it as a kind of shared safety backend for that wave of companies. And its move from general purpose moderation tooling toward AI specific defense, including red teaming and adversarial testing for model launches, reflects how the threat has changed. The same generative tools that created new products also handed attackers cheaper and faster ways to cause harm, and companies like Cinder are betting that platforms will keep paying for software that helps them keep up.