G42 gave up Huawei, then let Washington watch its data centres. Majority US ownership is reportedly next.

G42 executives have held exploratory talks about selling a majority stake to American companies to secure continued access to advanced chips beyond next year, Bloomberg reported. No deal terms or timing were disclosed, so the news is primarily about potential ownership/strategic repositioning rather than a confirmed transaction.

Analysis

This reads less like a classic control transaction and more like an export-control arbitrage: the valuable asset is not the equity itself, but the ability to keep buying frontier accelerators without a licensing interruption. If a U.S. shareholder structure is what unlocks that access, the first beneficiaries are the chip and networking vendors that sell into the AI buildout, because incremental demand can continue without a multi-quarter procurement pause. The market should focus on who captures the rent from the compliance workaround, not who owns the operating company.

The immediate risk is that this becomes a slow-moving political process rather than a clean catalyst. Over the next 1-3 months, the key events are not deal rumors but signals from Washington on what ownership/governance threshold is sufficient to preserve chip flow. If approval remains uncertain, the trade is basically a volatility cap on the ecosystem: AI capex in the Gulf gets delayed, which is a near-term negative for suppliers most levered to large cluster deployments, especially networking, storage, and power infrastructure.

Over 6-18 months, the structural implication is that export controls may increasingly force Middle East AI buyers into U.S.-aligned cap tables, creating a winner-take-most dynamic for American strategic capital and vendors already embedded in the stack. The contrarian view is that this may be overread as a bullish M&A story when it is really a defensive move to avoid losing access altogether; if the U.S. demands too much governance or disclosure, the transaction can collapse and the upside reverts to zero. What would falsify the bull case is any indication that licenses remain unchanged without a sale, or that regulators object to the buyer even after restructuring.

Key Decisions for Investors

• Keep a tactical long bias in NVDA and ANET for 1-3 months on the view that Gulf AI capex is preserved if U.S.-aligned ownership clears export friction; risk/reward is asymmetric because downside is limited if the deal stalls, but upside improves if licensing language becomes clearer.
• Use SMCI as a higher-beta expression only if procurement orders from the region reaccelerate; otherwise avoid chasing the stock on rumors because the trade depends on actual rack shipments, not governance headlines.
• For a relative-value trade, long semicap hardware (NVDA/ANET basket) vs. short a basket of non-U.S. AI infrastructure beneficiaries that are more exposed to export-control leakage risk; this captures the second-order winner from continued controlled access.
• Set an alert on any U.S. Treasury/Commerce Department commentary over the next 30-60 days; if regulators signal a hard line on ownership/board control, fade the rumor-driven rally in the AI supply chain.
• If a credible U.S. strategic buyer emerges, consider a 6-12 month long MSFT versus short broader EM AI exposure, since the structure would favor U.S.-tethered platform control over independent regional AI champions.