Crisis Text Line launched in 2013, growing out of the youth nonprofit DoSomething. Its idea was simple: anyone in crisis could text a number and reach a trained volunteer counselor, free, at any hour. By 2019 it had passed 100 million messages.
It was funded largely by big philanthropic gifts, including $23.8 million in 2016 from backers such as Reid Hoffman, Melinda Gates, the Ballmer Group and Omidyar Network. Revenue swung widely from year to year: $14.7 million in 2016, $3.4 million in 2017, $27.1 million in 2018.
Its conversations also made up one of the largest bodies of mental health text data anywhere. Crisis Text Line already shared scrubbed versions with outside researchers under formal agreements.
By late 2017, companies and a law enforcement agency had asked Crisis Text Line whether it could train their staff to handle hard conversations the way its counselors did. That was a real opportunity: a product, a market and possible income for a free service.
The board didn't want to run a training business inside the nonprofit and risk distracting from the crisis work. So it let the founder start a separate for-profit, Loris.ai. Crisis Text Line kept a significant stake and a board seat. Loris got controlled access to scrubbed conversation data, under an agreement modeled on the one for researchers, solely to build its training models. The board rejected any plan that gave Loris full access to its systems.
Board members had different reasons for saying yes. Some saw a way to reduce financial risk, since a stake that paid dividends or could be sold might help fund the crisis line. The year before Loris launched, revenue had dropped to $3.4 million. One founding board member, danah boyd, voted yes for another reason: if more people learned these skills, "perhaps the need for a crisis line would be reduced."
"No selling of data, period."danah boyd, founding board member, describing the board's rule
Loris.ai launched with a $2 million seed round, led by Floodgate, with Crisis Text Line's founder as its CEO. Crisis Text Line held a significant share, and Loris received scrubbed conversation data under the agreement.
Politico reported the arrangement. Critics focused on consent. Texters were shown a long privacy agreement, and Stanford privacy fellow Jennifer King told Politico: "These are people at their worst moments."
FCC Commissioner Brendan Carr wrote to both organizations asking them to "cease this data sharing and monetization practice," calling it "disturbingly dystopian." The same day, Crisis Text Line announced: "we have ended our data-sharing relationship with Loris." It noted that Loris "has not accessed any data since the beginning of 2020."
Crisis Text Line defended how the data had been handled. Its general counsel told Politico that "sensitive data from conversations is not commercialized, full stop," and the organization later said its practices were fully disclosed and lawful.
By then the founder had left the organization. It was led by CEO Dena Trujillo.
The payoff from the original bet was always speculative: a stake in a startup that might someday pay out. The risk was concentrated in the one thing a crisis line cannot work without, the confidence of the people who text it.
When that risk arrived, it arrived all at once. Critics on social media called it a "shocking betrayal of trust." In March 2022, Commissioner Carr asked the Federal Trade Commission to investigate Crisis Text Line's data practices, and a law firm invited affected users to come forward. Even boyd, who had voted for it, wrote that the board had concluded "we were wrong to share texter data with Loris.ai." Her verdict on the consent model: "A ToS is not consent."
Reversing cost little in practice, since Loris hadn't touched the data in two years. That is the hard part of the lesson. By the time the organization acted, the trust cost had already been paid.
Loris kept raising money. In April 2022 it raised $12 million, saying it uses only anonymized, aggregated data scrubbed of personal information. We found no public report that any texter was re-identified. No outcome of the FTC request has been made public.
Crisis Text Line updated its terms and privacy policy and promised to keep improving how it asks for consent. Dena Trujillo remained CEO through 2025, according to its tax filings. Its finances stayed volatile, as they had been before: revenue fell from $49.6 million in 2020 to $9.9 million in 2021, then climbed back to $45.4 million by 2024. Nothing in the record ties those swings to the Loris decision.
The debate outlived the deal. In 2024 a former volunteer counselor published a commentary in the Journal of Medical Internet Research arguing that agreeing to terms of service is not adequate research consent for people in crisis.
Guardrails protect the structure, not the trust.
The board built a separate company, scrubbed data and a no-selling rule. The public judged it on a different question: did the texters agree?
The people you serve define consent, not your terms of service.
A former board chair put it in five words: "A ToS is not consent."
A speculative payoff can carry a concentrated risk.
The upside was a stake that might pay someday. The downside sat in the one asset the service can't replace.
Fast reversal limits the damage but doesn't undo it.
Crisis Text Line acted in three days. The regulatory letters and critiques kept coming, and the consent debate was still in print in 2024.
If the people you serve saw exactly how their information, time or stories are being used, what would they say?
Where are you relying on a form someone clicked through, rather than an agreement they would recognize?
If you had to reverse a decision in three days, what would come back, and what wouldn't?
- danah boyd, "Crisis Text Line, from my perspective", apophenia, January 31, 2022. Board rationale, guardrails, significant share, "No selling of data, period," "we were wrong," "A ToS is not consent."
- Crisis Text Line, "An Update on Data Privacy, Our Community and Our Service", January 31, 2022. Ending the relationship, deletion request, no access since 2020.
- Commissioner Brendan Carr, letter to Crisis Text Line and Loris.ai, FCC, January 31, 2022.
- Megan Rose Dickey, "Loris.ai, a Crisis Text Line spin-out, raises $2 million", TechCrunch, February 6, 2018. Origin of the training request, seed round.
- Lonnie Lee Hood, Futurism report on Crisis Text Line data sharing, January 28, 2022. Summary of the Politico report, Jennifer King and general counsel quotes.
- Colleen Hagerty, "One of the biggest suicide prevention hotlines amends its data-sharing practices", Popular Science, February 1, 2022.
- Benton Institute, "FCC Commissioner Carr calls for FTC probe of Crisis Text Line", summarizing Politico, March 29, 2022. FTC request; Crisis Text Line's response that its practices were fully disclosed and lawful.
- The Conversation via Chronicle of Philanthropy, "Charities are contributing to growing mistrust of mental health text support", April 4, 2022. "Shocking betrayal of trust" as a description of social media reaction.
- Kyle Wiggers, "Fresh from controversy, call center analytics firm Loris raises $12M", TechCrunch, April 26, 2022.
- Timothy D. Reierson, commentary in the Journal of Medical Internet Research, December 30, 2024. Loris incorporation in November 2017; consent argument.
- Lieff Cabraser, "Crisis Text Line: Alleged Data Sharing Violations", undated.
- ProPublica, Crisis Text Line Inc. Nonprofit Explorer. Revenue 2016 to 2024; CEO listings.
- Wikipedia, "Crisis Text Line". Founding, 100 million messages, 2016 funding round.
An independent analysis of a publicly documented decision, written for leaders facing similar choices in their own work. For informational purposes only. No coaching relationship with the featured leaders, and no endorsement by them, is implied.