Events calendar back in question

Good afternoon friends, and happy hump day. The season’s events calendar is back in question in some corners, as the Red Sea Film Foundation canceled this year’s edition of its international film festival. The event will return to Historic Jeddah in 4Q 2027 at a yet-to-be-determined date. The announcement — which did not mention the ongoing war — attributed the decision to a review of the timing and extensive logistical needs required for an event of this scale.

In Dubai, some big events are also on pause as the emirate observes 10 days of mourning for Sheikh Ahmed bin Rashid Al Maktoum, brother of Dubai Ruler Sheikh Mohammed bin Rashid, who died on Monday at 75. Flags are at half-mast, live entertainment is off, and official gatherings are being pushed back through to the end of next week.

Up the highway: First Abu Dhabi Bank was named in a Financial Times investigation into how a Kremlin-backed fintech moved USD 6.9 bn past sanctions checks. A7 — a fintech player that provides cross-border payments, backed by the sanctioned Russian state bank Promsvyazbank — has been using forged invoices to defeat the sanctions compliance checks major banks run on their customers.

Who carried the transactions? Standard Chartered’s Hong Kong accounts alone received USD 1.1 bn from A7-linked entities; Citigroup, Deutsche Bank and JPMorgan also processed A7-linked payments, while First Abu Dhabi Bank held accounts for 17 A7 entities that sent over USD 1.8 bn outbound. Just over half the flows ultimately landed in China, with some payments traced to Russian military hardware and security-service purchases.

OpenAI wants Washington leading on AI rules days before Sam Altman briefs the UN Security Council on AI risk. In a Monday blog post, the company called on the US to lead an international push for technical standards on frontier AI, including recursive self-improvement (RSI) — the point where models start upgrading themselves. Fully autonomous RSI “is not happening today,” OpenAI said, and shouldn’t until it can be done safely.

Separately, the US is pitching to China a “notification mechanism” on AI security incidents ahead of the Trump-Xi summit tomorrow. However, sources familiar with the matter told Politico that the pitch is much more rudimentary than the name suggests, with one source describing it more as an open communication channel between US Treasury Secretary Bessent and China’s Vice Premier He Lifeng, rather than “a formal body of experts.”

IN CONTEXT- The US almost attacked a Chinese vessel a few months ago based on AI hallucinations. CNN reported earlier this week that the US came so close to intercepting a Chinese vessel in the region after an AI-generated intelligence report falsely flagged the vessel as transporting components for a nuclear program to an unidentified country.

There’s no reason to believe that a global “notification mechanism” for AI would emerge, given the intense geopolitical nature of the AI race between China and the US. And the global track record on similar efforts is dismal — several efforts to create a global notification mechanism on cybersecurity incidents have failed in delivering a system, leaving the effort to a set of domestic, bilateral, and multilateral frameworks along geopolitical lines. –Salma