'People Are Concerned' Over The Data Center Boom In Georgia Says Leavell

The article is a Bloomberg interview with Chuck Leavell discussing Georgia's upcoming primaries, AI, and the future of the music industry. It also touches on local concerns in Georgia, including the rise of data centers. The piece is largely conversational and informational, with limited immediate market implications.

Analysis

The investable angle is not the celebrity interview itself; it is the renewed political salience of data-center siting and the broader “AI infrastructure” backlash. That creates a two-sided setup: utilities, grid equipment, and gas-fired generation names can benefit if permitting and load growth stay intact, while hyperscalers and data-center REITs face a rising risk premium from local opposition, congestion charges, and longer interconnection timelines. The market often prices AI demand as if power is frictionless; the second-order reality is that megawatt availability, not chips, may become the binding constraint over the next 12-24 months.

The closest near-term catalyst is state-level policy scrutiny, which can move faster than federal AI regulation and disproportionately affect projects already in the pipeline. If Georgia becomes a template, the losers are not just new data-center developers but also adjacent beneficiaries of the buildout thesis: electrical contractors, transformer suppliers, and natural-gas midstream names that depend on load additions. Conversely, regulated utilities with constructive rate-base treatment and the ability to win cost recovery could see multiple expansion if investors conclude that AI load is more durable than feared.

Consensus is probably underestimating how much of the AI capex trade is really a power-trade in disguise. The market has been rewarding “picks and shovels” broadly, but the differentiated winners are the bottleneck owners: transmission, switchgear, and generation assets with long queue positions. The contrarian risk is that political resistance slows incremental demand enough to compress the expected revenue ramp for AI-adjacent infrastructure, even if headline AI spending remains strong. Over months, that argues for a more selective long basket and a hedge against over-owned data-center exposure.

Key Decisions for Investors

• Long XLU / short a basket of data-center REITs (e.g. EQIX, DLR) for 3-6 months: express the view that regulated utilities with rate-base growth are better insulated than capacity-dependent landlords; target a 10-15% relative move if permitting friction rises.
• Long infrastructure bottlenecks over headline AI beneficiaries: pair long ETN or HUBB against short NVDA on a 6-12 month horizon to capture power-grid spend without relying on multiple expansion in the most crowded AI name.
• Buy call spreads on utilities with heavy AI load exposure but constructive regulation, such as SO or DUK, dated 6-9 months out: asymmetric upside if markets start capitalizing durable load growth while limiting downside if political scrutiny delays projects.
• Short high-beta data-center development names on any rally tied to AI enthusiasm, using a 1-3 month horizon: the risk/reward favors fading optimism because interconnection delays and local opposition can push revenue recognition right while leaving capex upfront.