Unethical lab behavior includes any scenario where a researcher fabricates data, manipulates results, takes credit for others’ work, or deliberately ignores safety and ethical protocols. The U.S. Office of Research Integrity defines research misconduct as three core offenses: fabrication, falsification, and plagiarism. But unethical behavior in a lab extends well beyond that official triad into areas like authorship abuse, protocol violations, undisclosed conflicts of interest, and safety negligence.
Fabricating or Falsifying Data
Fabrication means making up data or results entirely and recording them as though they were real. A researcher who never ran an experiment but enters invented numbers into a spreadsheet is fabricating. Falsification is subtler: it involves manipulating materials, equipment, or processes, or changing and omitting data so the research record no longer reflects what actually happened. Both are clear-cut misconduct, and neither is considered an honest error or a difference of scientific opinion.
Common falsification scenarios include removing data points that contradict a hypothesis, adjusting instrument readings after the fact, or selectively reporting only the trials that produced a desired outcome. A particularly widespread form is known as p-hacking: manipulating how data are collected or analyzed until a result reaches statistical significance. This can look like stopping an experiment the moment a significant result appears rather than collecting the full planned sample, removing outliers without justification, or adding and dropping variables from a statistical model until the numbers cooperate. These practices inflate effect sizes and produce findings that look real on paper but don’t hold up.
Manipulating Images and Figures
Digital image manipulation is one of the most frequently flagged forms of misconduct in published research. Scientists are generally allowed to adjust brightness or contrast on lab images like western blots, but those adjustments must be applied uniformly across the entire image. When a researcher selectively lightens the background of a blot until it turns white, faint bands can disappear entirely, and splice marks become invisible. That crosses the line from acceptable enhancement into falsification.
Other problematic scenarios include cropping images so tightly that important context is removed, duplicating an image from one experiment and presenting it as data from a different experiment, or physically cutting membranes before imaging so that controls run on the same gel can’t be verified. Proper practice requires presenting images with a light gray background distinguishable from the white page, labeling any spliced sections, and retaining unedited source files.
Ghost and Guest Authorship
Authorship on a scientific paper is supposed to reflect who actually did the work. Ghost authorship occurs when someone who made a significant contribution, such as designing the study, running the analysis, or writing the manuscript, is left off the author list entirely. Guest authorship is the reverse: a person is named as an author despite not meeting authorship criteria, typically because their reputation lends credibility to the paper.
The most egregious version combines both. A pharmaceutical company hires professional writers and statisticians to design a study and draft a manuscript, then pays a well-known researcher to attach their name and submit it to a journal. The actual authors are hidden, the guest author didn’t do the work, and the journal and its readers have no way to evaluate the paper’s independence.
Violating Approved Research Protocols
Any research involving human participants or animals must follow a protocol approved by an oversight board. A protocol violation is any departure from that approved plan that affects participant safety or the integrity of the data. Specific scenarios that qualify as violations include enrolling a participant who doesn’t meet the study’s eligibility criteria, giving a subject the wrong treatment or incorrect dose, failing to withdraw a participant who met withdrawal criteria, and changing the protocol without prior approval.
The most serious category involves willful breaches: performing study procedures on a person before obtaining their informed consent, falsifying research or medical records, breaching participant confidentiality, or conducting procedures beyond one’s professional credentials. Repeated minor deviations also escalate into formal violations, even if no single instance seems severe on its own.
Hiding Financial Conflicts of Interest
A researcher who holds stock in a company sponsoring their study, consults for the manufacturer of a drug they’re testing, or receives personal compensation from an industry partner has a financial conflict of interest. The conflict itself isn’t automatically unethical. Failing to disclose it is.
Transparency is the primary tool for managing these situations. A conflicted investigator is typically required to disclose the relationship in informed consent documents when human subjects are involved, to journal editors when publishing results, to conference audiences when presenting findings, and to students on the research team whose academic careers could be affected. When a lab PI conceals equity in a sponsor, hides a consulting arrangement, or omits disclosure from a consent form, they compromise the integrity of the entire study and put participants and trainees at risk.
Ignoring Laboratory Safety Protocols
Deliberately disregarding safety procedures is a distinct category of unethical lab behavior. In microbiology labs, this can mean culturing unknown environmental organisms on an open bench without proper containment, bringing clinical isolates from hospital patients into a teaching lab not equipped for them, or allowing untrained staff to handle plates that could contain dangerous pathogens. One documented scenario involved a staff member unknowingly exposed to soil samples from an area historically endemic for anthrax, with no safety protocol in place for reporting or diagnosis.
The ethical obligation runs in two directions. Individual researchers are expected to follow updated biosafety guidelines and avoid methods that carry unnecessary risk. Institutions are expected to provide the biosafety officer support, equipment, and training that make safe practice possible. When either side cuts corners, whether out of habit, budget pressure, or a “let’s try it and see” attitude, students, staff, and the public face preventable harm.
Using AI to Generate Fake Data
Generative AI has introduced a new class of unethical scenarios. These systems can produce highly realistic microscopy images, radiological scans, clinical trial datasets, and epidemiological figures that are difficult to distinguish from real data. A researcher under pressure to publish could use AI to fabricate an entire dataset supporting their hypothesis, then present it as genuine experimental results in a paper or grant application.
AI tools can also be used to subtly alter real data, enhancing support for a preferred conclusion rather than generating results from scratch. Even when synthetic data are used legitimately (for example, to correct biases in a training set), failing to disclose, describe, and label them as synthetic creates ethical problems. Downstream researchers may unknowingly treat fabricated data as real, building flawed conclusions on a false foundation. Companies could also generate fake data to support regulatory applications for product approval, extending the risk well beyond academic publishing.
What Happens When Misconduct Is Reported
Lab members who witness unethical behavior can report it internally or, for federally funded research, to the Office of Research Integrity. Federal regulations require every institution receiving public health research funding to establish policies that protect the positions and reputations of people who report allegations in good faith. A whistleblower has 180 days from the date they became aware of retaliation to file a formal complaint, and if the institution doesn’t respond within ten working days, the complaint can go directly to ORI. Whistleblowers can also request interim protective actions at any point before their complaint is fully resolved.
Consequences for confirmed misconduct range from mandatory retractions and loss of funding to supervision requirements, debarment from federal grants, and termination. The Office of Research Integrity publishes case summaries of individuals with active administrative actions against them, making findings a matter of public record for the duration of the imposed sanctions.