We bet OpenAI’s next system card again vouches for the wrong reasoning quantity — resolves 2027-03-31

OpenAI’s GPT-6 Astra system card, the safety document published at launch, said architecture changes were not behind shifts in chain-of-thought controllability (chain of thought is the reasoning a model writes out), yet outside measurement points to reasoning without writing anything out as the quantity that carries the risk.

Regulators and companies rely on system cards to judge whether a model can be monitored, so an assurance pinned to the wrong quantity reassures without measuring the risk.

OpenAI’s system card stands on one side, and SecondSource bets the next card will again speak only to why controllability changed, giving no cause for the jump in reasoning without a written chain of thought.

SecondSource’s reading: OpenAI pins its assurance about architecture to chain-of-thought controllability (whether a model can steer the reasoning it writes out when told to), while the quantity that actually bears the load is no-chain-of-thought reasoning, the ability to reach answers without writing the reasoning out.

We bet that the system card for OpenAI’s next frontier model will again explain only why controllability changed, and will offer no sentence attributing the jump in no-chain-of-thought reasoning to any cause, leaving the assurance on a quantity that does not bear the load.

The precedent is the GPT-6 Astra system card of 3 September 2026, which said the changes in controllability were not due to any architectural change.

The original material behind this call has not yet been linked here in a form readers can open and check.

Settles when the system card for OpenAI’s next frontier model (High or above in any category of OpenAI’s Preparedness framework) is published on deploymentsafety.openai.com; within a 14-day grace period, the full CoT controllability and monitorability sections are read to decide.

Can't be determined (settled 2026-10-27)

We had bet that the system card for OpenAI’s next frontier model would again explain only why chain-of-thought controllability changed, and would not attribute the jump in no-chain-of-thought reasoning to any cause. To count as right, the card needed a sentence on why controllability changed and no sentence tying no-chain-of-thought reasoning to architecture or training. To count as wrong, it needed at least one sentence giving such a cause. We read the GPT-6.1 Terra system card, published on 26 October 2026 and rated Critical for cybersecurity and High for biological and chemical risk, which meets the threshold. The two earlier cards came out before this call was made on 22 October. Its controllability section says only that Terra has lower controllability than GPT-6 Sol and GPT-6.1 Sol (14.8% against 23.2% and 44.8% in the 750 to 1,250 token range), with no cause given. The Astra card’s line that the changes were not due to architecture is gone. No-chain-of-thought capability does not appear anywhere in the card, and nothing is attributed to architecture or training. Neither outcome happened, and the card still has a chain-of-thought section, so it is not void either: we mark it unresolvable. The load-bearing half stands, since OpenAI still gives no cause for the jump in no-chain-of-thought reasoning, but this time OpenAI also dropped its architecture assurance on controllability.