A base case is a most-likely estimate, which is a different object from a target. A company can forecast an overshoot of 1.5C and still hold a 1.5C-aligned commitment; the two documents answer different questions, and finding a gap between them is not by itself evidence of anything.
What makes Rio Tinto's disclosure worth reading closely is the point where the forecast stops being a view about the world and becomes an input. The company states that no portfolio adjustments are made to its future production plans [6], and its central case of 2.1C to 2.3C by 2100 rests on developing countries missing or delaying the net-zero targets they have already stated [5]. That is an assumption doing load-bearing work on capital allocation.
Set the numbers next to each other. The science summarised in the analysis puts the tolerable peak at 1.6C to 1.8C, which is 1.5C plus a limited overshoot of 0.1C to 0.3C, if warming is to be brought back down [11]. Rio Tinto's central case sits 0.3C to 0.7C above that ceiling: 2.1 minus 1.8 at the low end, 2.3 minus 1.6 at the high end [15]. AGL and Origin Energy, which say their strategies hold beyond 2.6C [8], are describing robustness at least 0.8C above the top of the same window [16].
Now the evidence's limits, because they are real. The analysis does not say how many reports were examined, referring to "several" of Australia's biggest energy and resource companies and to "most of these reports" [18]. That is a set of examples with no denominator, and it cannot support a claim about Australian corporates in general. Some firms publish no temperature figure at all; the authors read their disclosures as inconsistent with 1.5C [9], which is an inference from the surrounding text rather than a quoted number.
The thing a disclosed base case does not tell you is the probability weight behind it, the sensitivity of the strategy to a cooler or hotter path, or the counterfactual: what the production plan would have been under a different assumption. Disclosure captures the scenario, not the decision the scenario produced. Rio Tinto is the exception only because it volunteers that the decision was to change nothing [6].
The analysis describes what it is seeing as a serious mismatch, and argues that when companies frame warming as a matter for government policy they understate the effect of their own financial decisions [17]. The framing is an argument; the filings are data. What would make the data much stronger is a full sample of the regime's first-year reports with probability ranges attached, rather than single point estimates read off a handful of documents. The regime now generates that material annually, which turns a snapshot into a repeated measure.
