Tenant targeting: who to prioritize
Briefing guidance lists priority tenant targets as enterprise AI software firms, infrastructure‑layer startups (deployment, observability, security), chip/systems adjacencies and growth‑stage companies likely to scale hiring once product traction turns into revenue. That segmentation is tied to the larger point that not all 'AI' firms create the same near‑term office demand. (thehindu.com) (intellectia.ai)
Landlords chasing “AI tenants” are being told to narrow the list: the near-term office demand is coming from enterprise software, infrastructure startups and chip-adjacent firms, not every company using artificial intelligence. (cbre.com) The office case starts with who is actually leasing. CBRE said AI-related companies had leased more than 5 million square feet in the United States by May 2025, and tech’s share of U.S. office leasing rose to 18% in the first three quarters of 2025 from 14.2% in 2023. (cbre.com 1) (cbre.com 2) That split matters because different AI businesses need different real estate. JLL said technology companies are redirecting space toward infrastructure, talent and innovation, while 56% reduced space in the prior year and only about 10% of portfolios were lab and research space. (jll.com 1) (jll.com 2) Enterprise AI software firms tend to look most like traditional office users: they hire sales, customer success, product and engineering teams once products start generating recurring revenue. JLL said 73% of technology companies added collaboration space even as many shrank their footprints, a sign that headcount quality matters more than raw square footage. (jll.com) Infrastructure-layer startups — the companies that help deploy models, monitor them and secure them — sit closer to the spending surge around AI computing. JLL’s 2026 data center outlook said AI represented about one-quarter of global data center workloads in 2025 and could reach half by 2030, with inference becoming the main driver. (jll.com) Chip and systems adjacencies also show up in leasing screens because hardware booms spill into office hiring before they show up in broad employment data. JLL said AI is reshaping priorities from semiconductor development to software innovation, tying real estate decisions to where companies are building products rather than where they merely market themselves as AI players. (jll.com) Recent company moves help explain the filter. Reuters reported on April 10, 2026 that Meta expanded its CoreWeave relationship with a fresh $21 billion cloud-computing deal, and Reuters reported on April 6, 2026 that British officials were trying to attract Anthropic with proposals that included a London office expansion. (thehindu.com 1) (thehindu.com 2) Anthropic has also been adding physical offices in markets where usage is turning into business. Reuters reported on October 24, 2025 that Anthropic planned a South Korea office, and on March 2026 that it opened a Bengaluru office after saying India had become Claude’s largest market after the United States. (thehindu.com) (thehindu.com) CoreWeave shows the other side of the equation: infrastructure demand can be huge without creating classic office demand at the same scale. Intellectia, citing Nasdaq.com analysis, said CoreWeave’s backlog reached $55.6 billion in 2025 and revenue was projected at about $12 billion in 2026, but that growth is tied heavily to data centers, financing and specialized compute capacity. (intellectia.ai) (intellectia.ai) The practical screen is simple: prioritize AI companies that are adding office workers, not just graphics processing units. In this market, a startup with revenue traction and a hiring plan is a better office prospect than a company with an AI label and no reason to add desks. (cbre.com) (jll.com)
