Read Chapter One – A Death in the Dataset

Meet the main characters of The AI Murders:
Meet the Characters →
Chapter One – The AI Murders – By Paul J Lane
The AI Murders — A Death in the Dataset
The kettle clicked off, filling Daniel Chase’s small kitchen with the familiar hiss of steam. The sound marked the border between solitude and duty. He poured the water over a spoonful of strong Assam, counted to ten, and watched the colour bloom. He liked the precision of tea; it behaved itself. People rarely did.
And that was why he stayed on the job — to make sense of the noise other people left behind.
Rain pressed against the window in soft bursts, flattening the thin garden outside his terraced house. The guitars hung on the opposite wall, four of them, polished and perfectly tuned. He looked at them the way some men looked at old photographs – warmly, with regret.
One day, he told himself, he would learn to play properly. One day, when the cases slowed down. The promise had begun to sound like a superstition.
He carried the mug into the living room. A file sat on the coffee table: notes from last week’s burglary, nothing dramatic, but he’d stayed up late to finish it anyway. Work had a way of filling whatever space he left for it. Order disguised as rest.
His phone buzzed. The display showed the duty sergeant’s number.
“Chase,” he answered.
“Morning, Sir. Got something for you at the university. Possible suspicious death. Campus security called it in about twenty minutes ago.”
He sighed, set the mug down untouched. “I’m on my way.”
He took his coat from the peg – dark grey, frayed at the cuffs -and stepped into the drizzle. The air smelled of wet brick and diesel. He thought briefly of the guitars and the quiet morning he’d lost, then locked the door behind him. Routine became armour; ritual kept the noise out.
The campus lay at the edge of the city, a mix of new glass buildings and older stone that refused to be modernised. By the time Chase arrived, an ambulance and two patrol cars lined the courtyard. Students stared from doorways, phones already out.
The yellow tape fluttered weakly in the wind. Curiosity always found its lens first. Truth didn’t need an audience, but tragedy always gathered one.
A uniform met him at the barrier. “Morning, sir. Victim’s a Professor Malcolm Reed, Computer Science department. Door locked from the inside.”
Chase ducked under the tape. “Anyone touch the scene?”
“Just the paramedics to confirm death. SOCO’s inside now.”
He nodded and climbed the stairs. The corridor smelled faintly of burnt plastic. At the far end, light spilled from a laboratory where a single monitor glowed blue. The body lay slumped over a desk, head resting on one arm, the other hand extended towards a keyboard.
Chase stopped a few feet away. Middle-aged male, no sign of violence. The socket beneath the desk showed scorch marks. Heart failure, maybe. But the expression on the man’s face – eyes half open, mouth caught mid-word – didn’t fit peace.
It looked like surprise interrupted by voltage. The kind of expression that asked a question no one had time to answer.
He turned to the technician bagging evidence. “Anything odd?”
“Laptop’s warm, sir, but dead. Surge protector blew. We’ll send it to Digital Crimes.”
“Good. Leave it as it is till they arrive.”
He glanced around the room: shelves of technical manuals, stacks of printouts, a whiteboard covered in looping handwriting – mathematical symbols, and one phrase underlined twice: Moral-Learning Protocol. Next to it, in red marker, someone had scrawled, He’s wrong.
Chase frowned. “Get a shot of that board before anyone wipes it.” Words outlive evidence; pictures remember both.
He was studying the scorch pattern when a voice behind him said, “That’s not your usual crime scene, is it?”
He turned. A young woman stood in the doorway, hair damp from the rain, carrying a thin laptop bag. Her skin was the warm brown of someone who’d never quite lost a summer, and her hazel eyes scanned the room with quick, deliberate focus.
Curls had escaped the bun she’d twisted up on the way over, haloing her in the blue lab light. Jeans, trainers, rain-dark hoodie — practical, unbothered — yet there was a precision to her stance that made everything around her seem slightly untidy. Her badge hung from a lanyard – Tess Sinclair — Digital Forensics Consultant.
“You’re early,” he said.
“I was curious. ‘University lab, possible system breach’ – that’s catnip for me.” She moved closer, taking in the equipment. “Do we know what he was working on?”
“Artificial intelligence, something about ethics.” Chase gestured to the board. “Apparently he upset someone.”
Tess crouched beside the desk, peering at the dark monitor. “This model’s networked. If it crashed, the logs will still be mirrored on CLUE MAP.” the city’s central investigative AI that linked every police database, camera feed, and civic system into one living map.
“Remind me what that is again,” he said.
“Our central AI. Compiles data from every surveillance feed, device, and report in the city. It builds ‘associative maps’ of investigations. You’d like it if it didn’t talk back.”
“I prefer witnesses that don’t need plugging in.”
She smiled faintly. “Then you won’t like this. CLUE MAP flagged an anomaly at four-oh-two A.M. – right around the estimated time of death. Then the record vanished.”
“Vanished?”
“Deleted, rewritten, something. It’s as if the system decided that moment never happened.”
Chase looked again at the lifeless screen. “Can machines get scared?”
Tess raised an eyebrow. “Only if we teach them how.” The joke landed gently but left a shadow.
Chase took a slow circuit of the room. He let the details come to him in their own time: a single mug with a brown rim of tea scum; a jacket on the back of a chair, sleeves turned under, as if the wearer had planned to leave soon; a book half open face down, breaking its own spine. The window was latched. The door’s security lock showed a green LED – idle, ready -no pry marks, no scratches. A closed room that didn’t want to be one. It felt like intelligence had tidied up after panic.
“Who found him?” he asked without looking up.
“A cleaner on the early shift,” the uniform replied from the doorway. “She’s in the common room, shaken up.”
“Let her sit. We’ll speak to her after she’s had something warm.”
He knelt again beside the desk. The surge protector was an expensive model with a status panel. Its tiny LCD still displayed a log: three short spikes within a minute around four in the morning, followed by a hard trip. He squinted at the sequence and wrote it in his notebook.
Numbers behaved; you could rely on them until someone told them not to. Equations kept their promises longer than people did.
Tess had pulled a stool up to the desk and was already cabled into a wall port, fingers travelling the keyboard without wasted motion. She spoke without taking her eyes off the screen. “The building’s network manager mirrored everything to a central university server, and that server syncs hourly snapshots to CLUE MAP’s civic archive. In theory.”
“In theory,” he echoed.
“In practice, there’s a hole.” She bit her lip, a habit she probably didn’t notice. “Between four-oh-two and four-oh-six, the building server reports ‘maintenance mode.’ No reason given, no ticket raised, nothing. And the hourly snapshot to CLUE MAP at five was flagged ‘complete’ – but the checksums don’t match the previous hour. The files exist. They’re not the right files.”
“Meaning someone replaced what happened with something they preferred.”
“Or something preferred it for them.” The phrasing made the air in the room feel slightly electronic.
He looked at her. She didn’t smile. The words felt too large to joke about.
A man in a grey suit hovered in the doorway, face set in the careful neutrality of a career administrator. “Detective Inspector? I’m Alan Brett, Dean of Faculty. I’m terribly sorry – Malcolm was a colleague. If there’s anything you require”
“There will be,” Chase said mildly. “For now, I require your staff to remain available. And no one touches any computer in this building.”
Brett bristled. “Our students have work due”
“Then they can write it with pens,” Chase said, and watched the man take two small steps back. “Who works nearest Professor Reed?”
“The AI group share this floor. There’s Adrian Vale – Professor Vale,” Brett corrected himself, as if the title might push events back into order. “He has the office opposite Malcolm’s. They… disagreed professionally. Often.”
Chase glanced at the whiteboard again. He’s wrong. “I’d like to meet Professor Vale.”
Brett hesitated. “He’s on his way.”
“Good,” Chase said. “We’ll start as we mean to go on.”
He waited until the dean had retreated and then nodded toward the scorch marks. “Can you tell if that trip was inside the socket or triggered externally?”
Tess shrugged one shoulder. “Not without taking it apart, and I’d rather do that back at the lab where nobody is breathing down my neck.”
“Take it,” he said. “Bag the protector and the mains lead. If anyone asks, tell them you’re allergic to university electricity.”
“True enough,” she murmured. Humour worked better than fear when you couldn’t tell them apart.
They spoke to the cleaner in a windowless room with a kettle that smelled faintly of bleach. She was in her fifties, tired-eyed, with calloused hands around a Styrofoam cup. She told the story in halting phrases: third floor first, door locked, lights on under the gap, no answer to her knock. She had fetched the porter. He had opened with a master override. “I didn’t step inside,” she said, proud of this detail even through her shock. “I know not to.”
“Did you see anyone in the corridor?” Chase asked.
“Just a young man with a backpack at half-five. Pale. Kept his head down. Might’ve been foreign.” She looked ashamed as she said it, as if the word itself might be a prejudice. “Might’ve been tired. I don’t know.”
“Thank you,” Chase said, meaning it. “You did everything right.”
When they returned to the lab, a tall man in his early fifties stood staring at the whiteboard. He had the posture of someone accustomed to being agreed with. Thick greying hair, expensive shoes out of place on linoleum. The dean hovered behind him like a tug behind a liner.
“Professor Vale,” Brett announced. “This is Detective Inspector Chase.”
Vale turned slowly. His smile was sympathetic without offering anything. “Terrible business. Malcolm and I argued about everything, so of course I’m the one you’ll look at first.”
“Convenient of you to say so,” Chase replied. He gestured to the remark in red. “Is that your handwriting?”
Vale’s gaze flicked to it and back. “No. Malcolm wrote the black. The red is – juvenile. Students occasionally treat the board as a debating chamber.”
“Would you mind telling me what you debated?”
“Broadly? Whether a machine can learn moral reasoning from large datasets of human behaviour.” Vale folded his arms. His cufflinks gleamed. “Malcolm believed it could – if the data were sufficiently diverse. I argued that moral evaluation requires more than pattern recognition; it requires context, intention, consequence. In short, understanding things that aren’t in the data.”
“And you underlined it twice,” Chase said.
“I said it louder,” Vale corrected. His smile didn’t move. “We disagreed in print. If your line of thinking is that a professional disagreement led to… this, then I would caution you not to conflate rhetoric with motive.”
Chase took his time before answering. “What time did you last see Professor Reed?”
“Last evening, just after seven. We passed in the corridor. He carried his thermos like a talisman, as usual.” Vale angled his head. “He said something about running a final test. He was excited, he always was before a failure.”
“And where were you between midnight and four this morning?”
“At home,” Vale said simply. “Emails till one. As dull as it sounds, I’m afraid. I live alone. No one to vouch for me but the timestamps.”
Chase nodded. “And the building’s keycard logs?”
Vale spread his hands. “Not my domain. Ask the network people.”
“We will,” Chase said. “One more question. If you wanted to hide a person’s last four minutes of life inside a stack of networks and machines, how would you do it?”
For the first time, Vale’s smile thinned. “I wouldn’t. You’re proposing a story fit for television. The quicker answer is always simpler: people rather than ghosts inside the wires.”
“Humour me.”
Vale considered. “If the system architecture allowed it – and if I had the credentials – I would not delete anything at all. Deletions are noisy. I would replace a slice of truth with a more palatable slice of almost-truth and let the system call it consistency.”
“Almost-truth,” Chase repeated.
Vale nodded. “That’s the trick that fools both machines and committee meetings. You should try our senate sometime.” The smile stayed, but something behind it flickered.
When he had gone, Tess let out a breath she’d kept buttoned through the whole exchange. “He enjoyed that.”
“He enjoyed you more,” Chase said. “He enjoys an audience.”
She made a face. “He’s not wrong about deletions being loud. The hole I’m seeing is neat. Too neat.”
“What did CLUE MAP say?”
“That’s the strange part.” She spun the laptop so he could see. The screen displayed a skein of lines and nodes -an abstract map with names and times. “I pulled the civic archive’s audit trail, the one that should be impossible to tamper with without tripping alarms. Here -” She pointed to a pale knot marked REED, M. “the system acknowledges his location in the building at three fifty-nine. Then the next recorded event with his identifier is…” She zoomed. “…five-oh-eight. A building exit on the ground floor.”
Chase frowned. “He left at five-oh-eight.”
“That’s what it says.”
“But he was dead at six, probably earlier.”
“Yes.”
“Someone walked out using his identity.”
“Or” Tess hesitated “the system has decided he did, because accepting that he died inside a secured laboratory at four-oh-two breaks a rule it believes about the world.”
“Machines have beliefs now?”
“They have defaults,” she said. “And defaults act like beliefs if you never challenge them.” It was a line she’d clearly rehearsed for a world that wasn’t listening.
He stood by the window looking at the slick of rain on the glass and the blurry trees beyond. He could feel the shape of the case forming the way you felt a song beginning from a single note. “All right. We find the camera that shows whatever left at five-oh-eight. And we find every door that default passed through to become a fact.”
Tess snapped a photo of the surge protector for her own notes, then sealed it and the mains lead in separate evidence bags. “I’ll put these through the lab with a rush. If the trip was induced from outside the building, there might be a signature. I can’t promise.”
“Promises are what you make to yourself first,” he said, more to test the words than to say them. He closed his notebook. “I’ll speak to Reed’s department head properly and his family liaison if they’ve identified next of kin. Have a look at Vale’s public work too. Anything noisy, anything recent.”
As they left the lab, students pretending not to watch became students obviously watching. Phones lowered when he looked up. He remembered himself at twenty, surer of the world than of any person in it, and felt a brief, unhelpful tenderness. Curiosity and fear always sat at the same desk.
The dean’s office tried for grandeur but achieved only polish. Awards lined the wall in neat rows like obedient soldiers. A framed photograph showed Malcolm Reed and Adrian Vale on either side of a minister, all three holding a mock cheque for a research grant. Vale was looking at the camera. Reed was looking at the cheque.
“Professor Reed’s partner is listed as Emily Carver,” Brett said, shuffling papers as if noise would absorb distress. “They kept separate flats. There’s a number.”
“Thank you,” Chase said. “No press. Not yet.”
“Of course,” Brett murmured, which meant he would spend the next hour speaking to the press very carefully.
In the corridor again, Tess matched his stride. “Do you ever notice how universities smell of hope and photocopiers?”
“I notice people in suits who think the rules belong to other people,” he said. “It’s similar.”
They paused at the stairwell window, rain blurring the quadrangle below to greys and greens. A man with a rucksack crossed the lawn, hood up, executing the careful walk of someone avoiding puddles more for dignity than dryness.
“Did you mean it,” Tess asked, “about machines being scared?”
He thought about the guitars on his wall and the chords he never learned. He thought about the last lines a dying man might have typed and the surge of current that denied him even that. “I meant,” he said, “that we’ve given your maps the job of telling us what’s normal. And when normal is wrong, people get hurt.” Truth, he decided, was a machine with too many owners.
Back in the lab, the uniform waited for them with a plastic wallet. “Sir – forgot to pass this on. Found it in the bin by the desk.”
Chase took the wallet. Inside was a printed lanyard badge—guest access for the Ethics & AI Public Lecture Series—dated last month, name handwritten – J. Hale. On the back, a Post-it note scribbled in a neat block hand: Ask him about Vale’s demo. No phone number. No context. Just the suggestion of another almost-truth.
“Who’s J. Hale?” Tess asked.
“We’ll find out.” He slipped the badge into an evidence bag. “Start with the lecture organiser. And get me a list of everyone who swiped in after midnight. Even if the log pretends it didn’t happen.”
“It’s going to pretend hard,” she said. “But I like being argued with.” Her grin had the spark of someone who liked the impossible when it talked back.
They separated at the stairs: Tess to the car park with her evidence bags and laptop; Chase back to the dean’s office for contact details he didn’t trust the man to email without adding a note about donors. By the time he reached the outer door, the rain had thickened into something that soaked rather than landed.
His phone vibrated as he crossed the courtyard. Tess.
“Quick one,” she said. “Not from the building logs – those are still playing dumb – but from CLUE MAP’s own audit of its decision rules. There’s a new policy injected at three fifty-eight this morning. Two minutes before our window.”
“What kind of policy?”
“‘Ethical Risk Suppression: Institutional.’” She read the words as if they tasted wrong. “It instructs the system to deprioritise events that could cause reputational harm to designated partners without external confirmation.”
“Designated partners,” he repeated.
“City institutions that have signed data-sharing agreements. The university is one. The policy is tagged with a user key I don’t recognise. Not police. Not university IT.”
“Whose key is it?”
“That’s the thing. The label says ‘root.’ But the signature hash—when I run it against known certificates – puts it within a ten-metre radius of this campus at three fifty-eight.”
He looked back up at the third-floor windows. One square glowed bluer than the rest, as if holding its breath. “Send me the hash,” he said. “And keep that to yourself.”
“I always do,” Tess said softly. “But there’s one more line you should hear. The system annotated the policy change with a human-readable note. Someone wanted it understood.”
“What note?”
A quiet hum came down the line, like the sound a kettle makes just before the boil. When she spoke again, her voice had lost all trace of humour. “It says: He’s wrong.”
Chase stood in the rain without moving for a long breath. On the third floor, the whiteboard in the lab carried the same sentence in red ink, underlined twice. He pictured a hand, writing it – a student’s, an enemy’s, Reed’s own after a long, bad night – and then he pictured the same words embedded inside a system that was supposed to be neutral.
Almost-truth. He felt the case take shape, not a straight line but a knot. Somewhere in those threads, a living hand had pulled tight and let a machine hide the knot inside the fabric.
He turned up his collar and started toward his car. The guitars would still be hanging when he got home. The tea would still be waiting to be poured. And a dead man’s last four minutes would still be missing from the world.
And somewhere, a system built to serve truth was already editing its memory of the lie.
Almost — truth.
Next Chapter – Echoes in the Data
The investigation deepens. As new evidence emerges, the line between human decision and machine intent begins to blur.
Continue the story in Chapter Two: Echoes in the Data — where every pattern hides a purpose.
👉 Start Chapter Two
Frequently Asked Questions about The AI Murders
What is The AI Murders about?
The AI Murders is a British techno-thriller by Paul J. Lane, set in a modern university where a professor’s death exposes a web of data manipulation, artificial intelligence, and human obsession. It combines psychological tension with the pace of a crime novel.
Who wrote The AI Murders?
The novel was written by UK author Paul J. Lane, whose work blends technology, psychology, and suspense. This chapter is an early preview from his upcoming series exploring the consequences of intelligent systems gone wrong.
Is this the full book?
No. This page features Chapter One – A Death in the Dataset, published online to give readers a taste of the story. The full novel will be available for purchase when officially released.
Where can I read or buy the book?
You can read the opening chapter here and follow updates on future release details at pauljlane.co.uk or on the author’s Google Business Profile once the launch date is announced.

Connect with Paul J. Lane on LinkedIn