This question asks how far autobiographies can be trusted as a way of learning about well-known historical figures, given their bias, and when other sources are needed.
Question type: To what extent
An ETG General Paper original study guide to the 2024 A-Level GP Paper 1 essay on society & culture. Not affiliated with, or endorsed by, UCLES, Cambridge Assessment or SEAB. A study aid, not an official answer.
The claim assumes reliability is a single property a source either has or lacks, when an autobiography is reliable for some kinds of knowledge (subjective experience) and unreliable for others (objective record).
domain: reliability depends on what you are trying to learn, autobiographies are strong primary sources for inner perspective and weak ones for impartial fact.
How to approach it. Calibrate the reliability of autobiography as a historical source, distinguishing what it is reliable for from what it distorts, rather than declaring it simply trustworthy or not.
Autobiographies are highly reliable for learning a figure's perspective, their values, motives and felt experience, and highly unreliable for learning the objective record, because the same authorial control that gives unmatched access to the inner life also enables omission and self-justification, so reliability tracks what you are trying to learn.
The defender of autobiography argues it is the only source written by someone who was actually there, a primary document of events and rooms no historian could otherwise enter, so dismissing it as unreliable discards irreplaceable testimony. This carries real force, much of what we know about closed regimes and private deliberations survives only because a participant wrote it down. However, being the only witness does not make a witness impartial, and the most unique testimony is often the most interested, so the autobiography's value as a primary source is precisely what obliges the reader to corroborate it, not to trust it, which keeps reliability conditional on cross-checking.
Autobiographies are reliable evidence of how a figure wished to be understood and unreliable evidence of what actually happened, so they are indispensable to the historian and dangerous to the credulous, best read alongside the record, never instead of it.
Reliability is not a fixed property of autobiography but varies with conditions, the writer's motive, the proximity in time, the availability of corroboration and the type of figure, so the reliable answer specifies when an autobiography can be trusted rather than judging the genre as a whole.
A serious objection says the conditions-based view dodges the real point, that every autobiography, whatever its motive or corroboration, is written from a single interested vantage and cannot see itself from outside, so the genre as a category is inherently limited and the question can be answered about the category. There is something to this, no amount of sincerity escapes the single perspective. However, this is exactly why the conditions matter, the shared flaw sets a ceiling on reliability, but motive, corroboration and proximity determine where beneath that ceiling a given autobiography sits, so acknowledging the universal limit does not collapse the differences between a checkable contemporaneous diary and an uncorroborated legacy memoir, it makes assessing them the whole task.
An autobiography is as reliable as its conditions allow, trustworthy where motive is candid, the account is corroborated and the writer was close to events, and unreliable where fame, hindsight and self-interest go unchecked, so the verdict is never the genre's, always the individual case's.
Option A runs on the domain axis (autobiographies are reliable for inner perspective, unreliable for objective fact, because authorial control is both strength and flaw), while Option B runs on the conditions axis (reliability varies with motive, corroboration and proximity, so it is judged case by case). One divides by what you want to learn, the other by the circumstances of the source.

These outlines are free to read. Book a trial lesson to have your own GP essays marked against the band descriptors, the way ETG teaches General Paper from the exam backwards.