When AI Dunnit.

AI is being promoted as a tool to reduce human error in criminal investigations and healthcare but, I assert AI creates a serious harm by its very nature; AI cannot be held accountable and accountability is how we mere humans fix mistakes for fear that we will be humiliated, be disciplined, lose our jobs - none of this applies to AI who merrily trots along even when people are harmed. Further, the real benefit of accountability is not punishment but, rather, preventing the same mistakes in the future and how do we do that with AI?

Angela Lipps, a grandmother from Tennessee, was falsely identified by the facial recognition software (FRT) Clearview AI, as part of a bank fraud scheme in Fargo, North Dakota. Angela was living a quiet life, caring for her family when she was arrested, jailed first in Tennessee and then in Fargo for almost six months until she was released. By then she was traumatized and had lost her home. The Fargo police chief Zibolski said, “We’re happy to acknowledge when we make errors, and we’ve made a few in this case, for sure.” His happiness is unlikely to be shared by Angela, and the promise of an an 'overhaul' of its AI policy shouldn't hide the fact that no one was held responsible for the harm to Angela - a vague wave at AI is not the same as true accountability.

Angela's false arrest is not unique; there have been many documented false arrests. Harm from errors of false positive FRT, like in the case of Angela are one problem, but what about false negatives when a true criminal is let go - who knows how many times that has happened unless the are finally apprehended and an analysis is done showing FRT was inaccurate. Research also shows that AI is "more prone to false positive errors when applied to people of color."

Police officers are trusting algorithms that they did not create and, quite frankly, don't understand. When reasons for false positives come to light, such as low image resolution, officers can use this as a warning but, how low is too low and what about people who aren't white, when is FRT reliable? I obviously have no answers, only questions and a discomfort with people being harmed only to have people in power vaguely wave at an algorithm rather than holding someone responsible but, who can they hold responsible?

The use of facial recognition is growing not just because it *may* help correct errors (while certainly engaging in errors) but because it's a big money maker, so the answers of accountability matter:

"The global face recognition market was almost nine billion dollars in 2025, with projected growth to over 30 billion by 2034. Over a third of this market is in the U.S., but there is wide adoption of FRT around the world... Ten percent of U.S. police departments use FRT. The NYPD made 2,878 arrests resulting from FRT in the first five years of its use. The Metropolitan Police in London report 100 arrests using FRT in conjunction with mounted security cameras, including a suspect accused of kidnapping. Police in New Delhi used FRT to identify almost 3,000 missing children, and FRT has been used to identify refugee children who have been separated from their family. The National Center for Missing & Exploited Children (NCMEC) has used a tool called Spotlight, which makes use of FRT, to identify children who are victims of sex trafficking. In 2023, the FBI worked with NCMEC to identify or arrest 68 suspects of trafficking."

AI in healthcare is also big business, according to a 2025 report by Research Insights: "The global AI In Healthcare Market size is projected to be valued at USD 26.6 Billion in 2024 and reach USD 187.7 billion by 2030."

AI is used in many clinical tools and embedded in medical devices - it's the latter situation that gives rise to this story:

"In June 2022, a surgeon inserted a small balloon into Erin Ralph’s sinus cavity at a hospital in Fort Worth, Texas. According to a lawsuit filed by Ralph, Dr. Marc Dean was employing the TruDi Navigation System, which uses AI, to confirm the position of his instruments inside her head.

The procedure, known as a sinuplasty, is a minimally invasive technique to treat chronic sinusitis. A balloon is inflated to enlarge the sinus cavity opening, to allow better drainage and relieve inflammation.

But the TruDi system “misled and misdirected” Dean,.. A carotid artery – which supplies blood to the brain, face and neck – allegedly was injured, leading to a blood clot...After Ralph left the hospital, it became apparent that she had suffered a stroke. The mother of four returned and spent five days in intensive care [and] a section of her skull was removed “to allow her brain room to swell.” She finds it, "hard to walk without a brace and to get my left arm back working, again.”

Who is to blame?

Matt Baxter, Director, Professional Liability, states, “From an insurance standpoint, AI is not really changing the exposure, because the liability still stands with the healthcare professional,” Baxter said. “They still have the same responsibility, whether they are using AI or not, to make sure the information is correct.”

One group of researchers cited the concern that puts who is responsible in question, because AI is a “black box”, "with no way to understand the AI's algorithm. This is problematic because patients, physicians, and even designers, do not understand why or how a treatment recommendation is produced by AI technologies. … Due to the black box feature, medical AI systems might make incomprehensible mistakes."

So, the doctor who does not understand the algorithm is held responsible for AI mistakes and, worse, holding him/her liable does nothing to protect the next patient from this algorithm.

Mistakes are common so the question of responsibility is crucial: "A new study from researchers at Stanford and Harvard found that even today’s best artificial intelligence (AI) models make serious errors in a significant portion of medical cases … with the top-performing AI models producing 12 to 15 errors per 100 cases and the worst-performing models making mistakes in 40 out of 100 cases."

Would suing the AI company responsible make things safer? Maybe the loss of money would make them revisit their tech and pull those that aren't safe.

Whatever the answer, the question must be asked: when, not if, AI makes a mistake, how are the right people held accountable and what is being done to ensure the mistake doesn't happen again? 

There is a reason that AI in law enforcement and healthcare are big business: they are two of the largest institutions we have because civil society, in the Aristotelian sense, has been organized around collective survival where individuals can fulfill their potential. Derived from our empathy and ethics, our laws are designed to protect us as a society and healthcare is designed to protect us as individuals, so no wonder they are fodder for making big bucks. Do we want AI - that's devoid of empathy and ethics, causing harm without an ounce of remorse - seeping into the two institutions that we created to keep us safe or do we want a way to use our ethics, our humanity, to keep AI in check? 

A Tragedy

In two trials in 2023 and 2024, Lucy Letby, a 33-year-old nurse in the neonatal unit at the Countess of Chester Hospital in England, was found guilty of the murder or attempted murder of 14 babies in her care between June 2015 and June 2016 and sentenced to life in prison. 

Her murder trial lasted more than ten months and captivated the United Kingdom. The press called Lucy, "UK's most prolific child serial killer in modern times". The judge highlighted, "the cruelty and calculation" of her actions and a mother of an infant girl stated, "I don't think we will ever get over the fact that our daughter was tortured till she had no fight left in her, and everything she went through over her short life was deliberately done by someone who was supposed to protect her and help her come home, where she belonged."

Prologue: 

All tragedy invokes the question of what could have been done to stop it, but the prologue is Lucy's life before this tragedy and it's remarkably normal, with no indication that she was a danger to the tiny, premature babies she looked after. She appeared to have been a psychologically healthy and happy, with many close friends and a dedication to nursing. She wanted to be a nurse since she was a teenager, 

“She’d had a difficult birth herself, and she was very grateful for being alive to the nurses that would have helped save her life,” her friend Dawn Howe told the BBC. An only child, Letby grew up in Hereford, a city north of Bristol. In high school, she had a group of close friends who called themselves the “miss-match family”: they were dorky and liked to play games such as Cranium and Twister. Howe described Letby as the “most kind, gentle, soft friend.” Another friend said that she was “joyful and peaceful. Letby, who lived in staff housing on the hospital grounds, was twenty-five years old and had just finished a six-month course to become qualified in neonatal intensive care. She was one of only two junior nurses on the unit with that training. “We had massive staffing issues, where people were coming in and doing extra shifts,” a senior nurse on the unit said. “It was mainly Lucy that did a lot.” She was young, single, and saving to buy a house. That year, when a friend suggested that she take some time off, Letby texted her, “Work is always my priority.”

Act I: The Trial

The prosecutors, in seven of the murder or attempted murder charges, relied on an academic paper written in 1989 by Dr. Shoo Lee, one of Canada’s most renowned neonatologists, on a rare complication in newborns — pulmonary vascular air embolism — to argue that Ms. Letby had intentionally injected air into their veins.

At her trial, Lucy suffered from PTSD, was barely coherent and, despite denying that she murdered anyone, she was found guilty of the murder or attempted murder of 14 babies.

Act II: 

Dr. Lee had retired to a farm in Alberta in 2021 and only heard of the case when Lucy's lawyer emailed him in 2023. Dr. Lee agreed to help with Ms. Letby’s request for an appeal because the expert witness had misinterpreted his work, but the court ultimately denied her request, saying Dr. Lee’s testimony should have been introduced at trial.

Dr. Lee assembled a team of neonatal specialists to look into the case with the caveat that the panel’s review would be released even if they found Lucy guilty.

Fourteen specialists from around the world assessed the clinical evidence and found: In all cases, death or injury were due to natural causes or just bad medical care.

“There was no medical evidence to support malfeasance causing death or injury” in any of the babies that Ms. Letby was charged with harming.

“If there’s no malfeasance, there’s no murder. If there’s no murder, there’s no murderer,” Dr. Lee said, adding, “And if there’s no murderer, what is she doing in prison?”

Some of the hospital staff, the panel concluded, were caring for the most critically ill or premature babies in a unit that was only meant to treat babies with lesser needs. 

Act III: The Hospital

The neonatal unit at the Countess of Chester Hospital, run by the National Health Service in the west of England, was found by the Royal College of Paediatrics and Child Health to have inadequate nursing- and medical-staffing levels and the increased mortality rate in 2015 was not restricted to the neonatal unit.

Burkhard Schafer, a law professor at the University of Edinburgh who studies the intersection of law and science, said,

 “Looking for a responsible human—this is what the police are good at. What is not in the police’s remit is finding a systemic problem in an organization like the National Health Service, after decades of underfunding, where you have overworked people cutting little corners with very vulnerable babies who are already in a risk category. It is much more satisfying to say there was a bad person, there was a criminal, than to deal with the outcome of government policy.”

That last sentence warrants repeating: It is much more satisfying to say there was a bad person, there was a criminal, than to deal with the outcome of government policy. It was precisely the need to find a culprit that led to the failure to understand the real reasons that babies died. 

Act IV: Yet to be written

We have a nurse in jail for life for murders she didn't commit and a hospital woefully underfunded, that put babies lives at risk. There is no way to rectify things. Lucy, even if released, will be irrevocably damaged. The families who lost their babies will never get them back. The underfunding of hospitals remains unchanged.

It's all a damn tragedy.