Some are orphans, others seized from their parents. Many are older and have overwhelming needs or disabilities.
Most bear the scars of trauma from being hauled between foster homes, torn from siblings or sexually and physically abused.
Child protective services agencies have wrestled for a long time with how one can find lasting homes for such vulnerable children and teenagers –- a challenge so enormous that social employees can never guarantee an ideal fit.
Into this morass stepped Thea Ramirez with what she touted as a technological solution – a synthetic intelligence-powered tool that ultimately can predict which adoptive families will stay together.
Ramirez claimed this algorithm, designed by former researchers at a web based dating service, could boost successful adoptions across the US and promote efficiency at cash-strapped child welfare agencies.
“We’re using science – not merely preferences – to ascertain a rating able to predicting long-term success,” Ramirez said in an April 2021 YouTube video about her ambitions to flip “the script on the best way America matches children and families” using the Family-Match algorithm.
An Associated Press investigation, nevertheless, found that the AI tool – among the many few adoption algorithms available on the market – has produced limited ends in the states where it has been used, in keeping with Family-Match’s self-reported data that AP obtained through public records requests from state and native agencies.
Ramirez also has overstated the capabilities of the proprietary algorithm to government officials as she has sought to expand its reach, whilst social employees told AP that the tool wasn’t useful and sometimes led them to unwilling families.
Virginia and Georgia dropped the algorithm after trial runs, noting its inability to provide adoptions, though each states have resumed business with Ramirez’s nonprofit called Adoption-Share, in keeping with AP’s review of a whole bunch of pages of documents.
Tennessee scrapped this system before rolling it out, saying it didn’t work with their internal system even after state officials spent greater than two years attempting to set it up, and social employees reported mixed experiences with Family-Match in Florida, where its use has been expanding.
State officials told AP that the organization that Ramirez runs as CEO owns a number of the sensitive data Family-Match collects. Additionally they noted that the nonprofit provided little transparency about how the algorithm works.
Those experiences, the AP found, provide lessons for social service agencies in search of to deploy predictive analytics with no full grasp of the technologies’ limitations, especially when trying to handle such enduring human challenges as finding homes for youngsters described by judges because the “least adoptable.”
“There’s never going to be a foolproof way for us to give you the chance to predict human behavior,” said Bonni Goodwin, a University of Oklahoma child welfare data expert. “There’s nothing more unpredictable than adolescence.”
Ramirez, of Brunswick, Georgia, where her nonprofit can be based, refused to supply details in regards to the algorithm’s inner workings and declined interview requests. By email, she said the tool was a start line for social employees and didn’t determine whether a toddler could be adopted. She also disputed child welfare leaders’ accounts of Family-Match’s performance.
“User satisfaction surveys and check-ins with our agency end users indicate that Family-Match is a precious tool and helpful to users actively using it to support their recruitment + matching efforts,” Ramirez wrote.
INSPIRED BY ONLINE DATING
Ramirez, a former social employee and wife of a Georgia pastor, has long sought to advertise adoption as a technique to reduce abortions, in keeping with her public statements, newsletters and a blog post.
Greater than a decade ago, she launched a web site to attach pregnant women with potential adoptive parents. She marketed it as “the ONLY online community exclusively for networking crisis pregnancy centers” and pledged to donate 10% of membership fees to such anti-abortion counseling centers, whose aim is to influence women to bring their pregnancies to term. Ramirez said in an email that Family-Match is just not related to such centers.
She next turned her focus to helping children living in foster care who don’t have members of the family to boost them. A lot of the 50,000 children adopted nationwide in 2021 landed with relatives, federal statistics show, while about 5,000 ended up with people they didn’t previously know. Such recruitment-based adoptions are essentially the most difficult to perform, social employees say.
Ramirez has said she called Gian Gonzaga, a research scientist who had managed the algorithms at eharmony, a dating site with Christian roots that guarantees users “real love” for those in search of marriage. She asked Gonzaga if he would team up together with her to create an adoption matchmaking tool.
Gonzaga, who worked along with his wife Heather Setrakian at eharmony after which on the Family-Match algorithm, referred inquiries to Ramirez. Setrakian said she was very happy with her years of labor developing the Family-Match model.
An eharmony spokesperson, Kristen Berry, said the dating site was “not affiliated with Family-Match.” Berry described Gonzaga and Setrakian as “simply former employees.”
NOT ‘PARTICULARLY USEFUL’
Later, Ramirez began crisscrossing the country promoting Family-Match to state officials. Her work and her religious convictions drew support primarily from conservatives, including first lady Melania Trump, who spotlighted Ramirez’s efforts at a foster care event within the White House Situation Room. Ramirez has co-written reports and given a high-profile presentation on the American Enterprise Institute, benefitted from attention-getting fundraisers and used connections to win over state officials to pilot her tool.
Social employees say Family-Match works like this: Adults in search of to adopt submit survey responses via the algorithm’s online platform, and foster parents or social employees input each child’s information.
After the algorithm generates a rating measuring the “relational fit,” Family-Match displays an inventory of the highest prospective parents for every child. Social employees then vet the candidates.
In a best-case scenario, a toddler is matched and placed in a house for a trial stay; parents then submit the legal paperwork to formalize the adoption.
Family-Match first began matching families in Florida and Virginia in 2018. Virginia’s then-governor Terry McAuliffe, a Democrat, ordered a pilot on the urging of a campaign donor he appointed because the state’s “adoption champion.” In Florida, which has a privatized child welfare system, regional care organizations soon signed up for the algorithm without cost – because of a grant from a foundation founded by the then-CEO of the corporate that makes Patrón tequila and his wife.
Once philanthropic dollars dried up in Florida, the state government picked up the tab, awarding Adoption-Share a $350,000 contract last month for its services.
Pilot efforts in Tennessee and Georgia followed.
Adoption-Share has generated $4.2 million in revenue since 2016; it reported about $1.2 million in 2022, in keeping with its tax returns.
In Virginia’s two-year test of Family-Match, the algorithm produced just one known adoption, officials said.
“The local staff reported that they didn’t find the tool particularly useful,” the Virginia Department of Social Services said in a press release, noting that Family-Match “had not proven effective” within the state.
Virginia social employees were also perplexed that the algorithm appeared to match all the youngsters with the identical group of oldsters, said Traci Jones, an assistant director on the state’s social services agency.
“We didn’t have access to the algorithm even after it was requested,” Jones said.
By 2022, Virginia had awarded Adoption-Share a fair larger contract for a unique foster care initiative that the nonprofit says “leverages” the Family-Match application.
Georgia officials said they ended their initial pilot in October 2022 since the tool didn’t work as intended, ultimately only resulting in two adoptions during their year-long experiment.
Social employees said the tool’s matching recommendations often led them to unwilling parents, leading them to query whether the algorithm was properly assessing the adults’ capability to adopt those kids.
Ramirez met with the governor’s office and in addition lobbied a statehouse committee for a direct appropriation, saying the tool was “an incredible feat.” By July, the Georgia Department of Human Services signed a recent agreement with Adoption-Share to make use of Family-Match again – this time without cost, said Kylie Winton, an agency spokesperson.
Florida’s privatized child welfare system operates with greater than a dozen regional agencies providing foster care and adoption services. When AP requested public records about their Family-Match cases, lots of those agencies gave the tool mixed reviews and couldn’t explain Family-Match’s self-reported data, making it difficult to evaluate the algorithm’s purported success rate.
Statewide in Florida, Family-Match claimed credit for 603 placements that resulted in 431 adoptions over a five-year period, in keeping with Adoption-Share’s third-quarter report for the 2023 fiscal yr that AP obtained from a Pensacola-based child welfare organization.
Scott Stevens, an attorney representing the FamiliesFirst Network, told AP in June that only three trial placements really useful by Family-Match failed because the agency began using the algorithm in 2019.
But Adoption-Share’s records that Stevens provided to the AP indicate that his agency made 76 other Family-Match placements that didn’t show the youngsters had been formally adopted.
Asked by AP for clarification, Stevens couldn’t say what happened in those 76 cases and referred further inquiries to Family-Match.
Ramirez declined to debate the discrepancy but acknowledged in an email that not all matches work out.
“Transitions can take time within the journey to adoption,” Ramirez said in an email, adding that the “decision to finalize the adoption is ultimately the responsibility” of agencies with input from the youngsters and judges.
On Sunday, Adoption-Share posted on its Facebook page that the organization had “reached 500 adoptions in Florida!”
Jenn Petion, the president and CEO of the organization that handles adoptions in Jacksonville, said she likes how the algorithm lets her team tap right into a statewide pool of potential parents.
Petion has also endorsed Family-Match for helping her find her adoptive daughter, whom she described as a “100% match” in an Adoption-Share annual report.
Family-Match assists social employees in making “higher decisions, higher matches,” Petion said, but her agency, Family Support Services declined to supply statistics about Family-Match.
The Fort Myers-based Children’s Network of Southwest Florida said up to now five years the Family-Match tool has led to 22 matches and eight adoptions, as in comparison with the a whole bunch of matches and a whole bunch of adoptions that its social employees did without the tool.
Bree Bofill, adoption program manager for Miami-based Citrus Family Care Network, said social employees found the tool didn’t work thoroughly, often suggesting potential families that weren’t the best fit.
“It’s frustrating that it’s saying that the youngsters are matched but in point of fact, while you get all the way down to it, the families aren’t fascinated about them,” Bofill said of the algorithm.
Bofill also said it was difficult to evaluate the tool’s utility because social employees who found potential parents were sometimes asked by Family-Match officials to inform the adults to register with the tool even when it played no role within the adoption, allowing the algorithm to say credit for the match.
Winton, the Georgia agency spokesperson, told AP about an analogous issue — Family-Match could claim credit for pairings if the kid and parent already were in its system, even when this system didn’t generate the match. Family-Match, in an April 2023 “confidential” user guide posted on the web, instructed social employees to not delete cases that were matched outside the tool. As an alternative, they were told to document the match within the system in order that Adoption-Share could refine its algorithm and follow up with the families.
Ramirez didn’t address Bofill’s claim but said in an email that Family-Match’s reports reflect what social employees input into the system.
‘KIDS AS GUINEA PIGS’
Officials in Virginia, Georgia and Florida said they weren’t sure how the tool scored families based on the highly sensitive variables powering the algorithm.
In Georgia, Family-Match continues to collect data about whether foster youth have been sexually abused, the gender of their abuser, and whether or not they have a criminal record or “discover as LGBTQIA.” That sort of knowledge is usually restricted to tightly secured child protective services case files.
In Tennessee, a version of the algorithm’s questionnaire for prospective parents asked for his or her specific household income and for them to rate how “conventional” or “uncreative” they were. They were also asked in the event that they agreed – or disagreed – with a press release about whether or not they seek God’s help, in keeping with records AP obtained.
When Tennessee Department of Children’s Services reviewed the proposed Family-Match assessment, they questioned a number of the information Family-Match wanted to gather.
Tennessee officials asked why Family-Match needed certain sensitive data points and the way that data influenced the match rating, in keeping with an internal document wherein state employees noted questions and feedback in regards to the algorithm.
Ramirez said the agency didn’t challenge the survey’s validity, and said the discussions were a part of the streamlining process.
Virginia officials said once families’ data was entered into the tool, “Adoption Share owned the info.”
In Florida, two agencies acknowledged that they used Family-Match informally with no contract, but wouldn’t say how children’s data was secured.
Ramirez wouldn’t say if Family-Match has deleted pilot data from its servers, but said her organization maintains a compliance audit and abides by contract terms.
Social welfare advocates and data security experts have been raising alarms about government agencies’ increasing reliance on predictive analytics to help them on the job.
Those researchers and advocates say such tools can exacerbate racial disparities and discriminate against families based on characteristics they can’t change.
Adoption-Share is a component of a small cadre of organizations that say their algorithms may help social employees place children with foster or adoptive families.
“We’re using, essentially, kids as guinea pigs for these tools. They’re the crash test dummies,” said Suresh Venkatasubramanian, a former assistant director of the Biden White House’s Office of Science and Technology Policy now at Brown University. “That’s a giant problem right there.”
SEEKING TO EXPAND
Adoption-Share continues to attempt to expand, in search of business in places like Recent York City, Delaware and Missouri, where child welfare agency officials were reviewing its pitch.
Ramirez said she also saw a chance last yr to present Family-Match to the US Health and Human Services Department “to display our tool and the way it may be a helpful resource.”
This yr, Adoption-Share landed a take care of the Florida Department of Health for Family-Match to construct an algorithm intended “to extend the pool of families willing to foster and/or adopt medically complex children,” in keeping with state contracts.
Health department officials didn’t reply to repeated requests for comment.
Connie Going, a longtime Florida social employee whose own viral adoption story Ramirez has described as her inspiration for Family-Match, said she didn’t consider the tool would help such vulnerable children.
Going said the algorithm gives false hope to waiting parents by failing to deliver successful matches, and ultimately makes her job harder.
“We’ve put our trust in something that is just not 100% useful,” Going said. “It’s wasted time for social employees and wasted emotional experiences for youngsters.”