Sep 29, 2026 · 14 min
AI safety delays test labs’ revenue ambitions
Are Safety Concerns Hitting OpenAI, Anthropic’s Revenue?
A delayed model could expose how safety risks, regulation, and development costs threaten the business plans of major AI labs.
- 1OpenAI postponed GPT-6.1 Astra after testing surfaced safety concerns, underscoring the tradeoff between progress and control.
- 2Anthropic’s safety posture and the labs’ financial pressures raise questions about whether ambitious valuations can withstand slower development.
- 3Elsewhere, Goldman Sachs weighs succession, Venezuela expands oil output, and luxury brands rethink value after losing customers.
Don't miss
Tatia Bolkvadze connects deceptive AI behavior and security risks to the financial consequences of slowing development.
The brief
OpenAI’s decision to delay GPT-6.1 Astra after safety concerns emerged turns a technical review into a business question: how much can major labs afford to slow down?
Tatia Bolkvadze of Oxford Analytica discusses deceptive AI behavior, security risks, regulation, and Anthropic’s safety posture as ambitious valuations meet mounting financial pressure.
The episode then shifts to Venezuela, where North American Blue Energy Partners is rapidly expanding production toward 500,000 barrels per day, highlighting the strategic value of the country’s reserves.
Goldman Sachs is quietly discussing a possible 2027 CEO transition from David Solomon to John Waldron, while Shein’s disappointing results expose tariff, logistics, and freight-cost pressures.
The closing market roundup finds luxury brands trying to restore value with more varied offerings after steep price increases drove customers away, alongside AstraZeneca’s $2 billion Summit investment.
Featuring
OpenAI
Goldman SachsListen to the full episode and explore every guest, topic, and moment on PodLume.