In a July 17, 2026, op-ed, for­mer Barack Obama advi­sor Van Jones and for­mer George W. Bush appointee David Safavian joint­ly argue that courts con­tin­ue to accept dis­cred­it­ed foren­sic tech­niques as reli­able evi­dence, con­tribut­ing to wrong­ful con­vic­tions across the coun­try. The authors con­trast the con­fi­dent foren­sic experts seen on tele­vi­sion crime dra­mas with real-world court­rooms where they say, pho­ny foren­sics” meth­ods—micro­scop­ic hair com­par­i­son, bite mark analy­sis, and blood-stain pat­tern inter­pre­ta­tion — are too often pre­sent­ed to juries as so-called evidence.”

To work, our jus­tice sys­tem should depend on reli­able, val­i­dat­ed evi­dence rather than unproven the­o­ries and junk sci­ence. Wrongful con­vic­tions rob inno­cent peo­ple of free­dom, drain mil­lions in tax­pay­er dol­lars through pro­tract­ed lit­i­ga­tion and unnec­es­sary incar­cer­a­tion, and force the gov­ern­ment to pay for the sub­stan­tial set­tle­ments required to redress when those con­vic­tions are overturned.”

Van Jones and David Safavian on the use of junk sci­ence in criminal cases.

The op-ed high­lights the case of Nick Flannery, an Ohio father charged with assault and child endan­ger­ment after his infant son was hos­pi­tal­ized with brain bleed­ing. Medical experts who lat­er reviewed Mr. Flannery’s case point­ed to a pre­ex­ist­ing con­di­tion rather than abuse, yet pros­e­cu­tors intend to pro­ceed to tri­al rely­ing on a Shaken Baby Syndrome diag­no­sis, a the­o­ry the authors con­sid­er long-dis­put­ed.” 

The authors also point to the case of Robert Roberson, a Texas death row pris­on­er who was con­vict­ed and sen­tenced to death for the death of his young daugh­ter based on Shaken Baby Syndrome evi­dence. Mr. Roberson’s exe­cu­tion was most recent­ly halt­ed in 2025 by the Texas Court of Criminal Appeals under Article 11.073, the state’s law that allows con­vic­tions to be recon­sid­ered when under­ly­ing foren­sic evi­dence is out­dat­ed or sci­en­tif­i­cal­ly dis­cred­it­ed. Mr. Jones and Mr. Safavian note that the lead detec­tive in Mr. Roberson’s case is expect­ed to tes­ti­fy about the unre­li­a­bil­i­ty of the evi­dence if Mr. Roberson receives a new tri­al. They also note that,because of the junk sci­ence in Mr. Roberson’s case, he came with­in a week of exe­cu­tion and “[t]hat’s how close the jus­tice sys­tem came to an irre­versible mis­take.” The authors also ref­er­ence Josh Burns (MI) and Andrew Roark (TX) as men whose con­vic­tions in sim­i­lar cas­es were lat­er over­turned on appeals. 

Citing Kentucky Innocence Project data find­ing that junk sci­ence con­tributed to a major­i­ty of the wrong­ful con­vic­tions it has reviewed, and to rough­ly a quar­ter of wrong­ful con­vic­tions nation­al­ly, the authors call on state courts to adopt stan­dards mir­ror­ing Federal Rule of Evidence 702, which requires foren­sic tes­ti­mo­ny to rest on reli­able sci­en­tif­ic prin­ci­ples and meth­ods. Mr. Jones and Mr. Safavian note that “[i]n Law & Order,’ the right per­son is usu­al­ly found guilty. In real life, that’s not always the case. If we are to have true jus­tice, judges must be empow­ered to set a high bar to screen out junk sci­ence and pro­tect the inno­cent.” They argue this would let judges screen out unre­li­able foren­sic tes­ti­mo­ny before it reach­es jurors and frame the reform as essen­tial to pre­serv­ing pub­lic con­fi­dence in the justice system.

Citation Guide
Sources

Van Jones and David Safavian, Junk sci­ence dri­ves wrong­ful con­vic­tions. That’s not jus­tice., USA Today, July 172026.