Weasel words / hedging
Vague qualifiers like 'may help' or 'studies show' that let a claim dodge commitment and leave the sentence saying nothing.
See it
What it is
Weasel words are qualifiers that look like a claim but survive any challenge: 'may help', 'up to 40%', 'studies show', 'one of the leading', 'virtually', 'arguably', 'helps support'. The name comes from the folk belief that weasels suck eggs and leave the shell intact, which is exactly what these words do to a sentence. Wikipedia's own manual of style bans them under that name.
Hedging is the broader habit: softening with 'somewhat', 'I think', 'it seems', 'quite', 'in many cases'. Some hedging is honest, because not every claim is certain and overclaiming is worse. Legal and medical copy often has to hedge. The difference is whether the qualifier carries information ('in trials of 200 users') or just absorbs risk ('could potentially help improve').
Gotcha: hedges cluster. One 'may' is caution, four in a paragraph reads as either lawyer-fear or AI filler, and readers now clock the pattern instantly. The fix is not deleting the hedge, it is finding the number, the source, or the scope that replaces it. If you cannot find one, the honest move is to cut the claim instead of blurring it.
Ask AI for it
Audit the copy below for weasel words and hedging. Find every qualifier ('may', 'could', 'up to', 'helps support', 'virtually', 'studies show', 'one of the leading', 'somewhat', 'arguably') plus every unsourced passive attribution, and sort each into one of four buckets: evidence-carrying (it names a real scope, source, or condition), uncertainty-calibrating (the outcome genuinely varies and the hedge is honest), legally required, or empty. Leave the first three alone and say why each one earns its place. For every empty hedge, return the original sentence and a rewrite that replaces it with a source, a scope, or a measurable claim. If no evidence exists, cut the claim rather than making it sound certain. Then give me the full rewritten copy at the end.