A new Barclays survey of 410 fixed-income investors reveals that AI has moved from testing to daily use in research and risk management, though humans retain final decision authority. Among long-only managers and asset owners, 52% use AI primarily for research, while 44% of hedge funds rely on it for market data processing. Hedge funds lead adoption with 72% using AI daily, versus 49% of long-only managers and 38% of asset owners. AI remains limited in trading and execution, with data security cited as the top barrier. Only 7% expect significant job cuts; most anticipate higher output. Separately, Marc Andreessen warns that AI's growth is constrained by energy and cooling infrastructure. The IEA projects data center electricity demand will more than double by 2030 to 945 TWh, near Japan's total consumption. In the US, data centers may soon consume more power than aluminum, steel, and cement production combined. Combined, the survey shows institutional AI demand is real, while Andreessen highlights the physical limits that will determine winners. Hyperscalers like Microsoft, Amazon, Alphabet, and Meta have $725 billion in combined 2026 capital guidance, up 77% from 2025, indicating the scale of infrastructure investment needed.

Key facts

• 52% of long-only managers use AI primarily for research.
• 72% of hedge funds report daily AI usage.
• Data security is the top barrier to wider AI adoption.
• Only 7% of investors expect significant job cuts from AI.
• Andreessen says AI growth depends on energy and cooling infrastructure.