Misinformation
Also called Misinformation
False information spread without intent to deceive — people sharing wrong things they believe are true.
Think of it like
A game of telephone where the message gets garbled honestly — nobody’s lying, but what comes out the end is wrong.
Example
A model confidently states an outdated medical guideline, a user believes it, and repeats it to friends as fact — no malice, real spread of a falsehood.
How it actually works
Misinformation is about error, not intent — the distinction from disinformation, which is deliberate. AI amplifies it two ways: models can generate confident falsehoods (hallucinations) that users trust and pass on, and they let anyone produce plausible content at scale. The defenses are the boring, hard ones: grounding outputs in sources, citing, fact-checking, and building systems that admit uncertainty instead of bluffing.
For product teams
A reputational and real-world-harm risk whenever your model states facts users will act on.
For engineers
Unintentional falsehoods, amplified by confident hallucination and cheap generation; mitigated with grounding, citation, and calibrated uncertainty.
Related
- Disinformation — The deliberate version.
- Hallucination — A key source of AI-generated falsehoods.
- Fact-Checking — The verification defense.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome