Commercial
CBRE predicts AI will cut U.S. office vacancy from 18.3% to 14.5% by 2031
By Real Estate Wire Staff, . Real Estate Wire.
CBRE is making a contrarian call on the office market. In a new analysis reported by Scotsman Guide, the commercial real estate services firm projects that artificial intelligence will be a net positive for office demand, not the sector's undoing.
The numbers CBRE puts forward are specific. According to Scotsman Guide's coverage, the firm estimates that only 5% of U.S. office jobs are highly vulnerable to AI disruption. Another 18% will likely endure or thrive, and the remaining 77% will see change but not significant displacement. On the back of that job picture, CBRE forecasts the national office vacancy rate declining from 18.3% in 2026 to 14.5% in 2031, with the office workforce growing at an annual rate of 0.9% over the next five years, outpacing projected overall U.S. job growth of 0.6% annually.
CBRE's case leans on two arguments. First, many of the jobs most exposed to AI, such as payroll clerks, tax collectors and revenue agents, are already performed remotely, so their elimination would not reduce physical office headcount in any meaningful way. Second, roles like financial analysts, computer network architects and AI engineers are expected to remain or become office-based positions, supported rather than replaced by the technology.
The historical analogy is central to CBRE's framing. Scotsman Guide quotes John Morris, CBRE's group president of advisory leasing: "History has shown that technological advancements often lead to more jobs rather than fewer. After the advent of both the internet and the smartphone, office-using jobs made large gains in share of overall U.S. jobs."
As a leading indicator, CBRE points to U.S. Census Bureau data showing new business applications rising from just over 400,000 per year in 2022, the year ChatGPT launched, to more than 500,000 in 2025.
The supply-side implication is where the forecast gets operationally interesting for investors and leasing teams. Mike Watts, CBRE Americas' president of office investor leasing, is quoted by Scotsman Guide saying that tightening availability in prime buildings will eventually push demand into the next tier down: "When there are no large blocks of available space in prime, best-of-the-best buildings in a given market, that's when the next tier of office buildings below prime likely will attract spillover demand from relocating occupiers."
CBRE also notes that minimal new office construction is baked into the forecast, which would accelerate the vacancy compression if demand materializes as projected.
It is worth being clear about what this report does not establish. CBRE is a major office leasing broker with a direct commercial interest in a bullish office narrative, and Scotsman Guide's coverage does not include any independent validation of the job-disruption estimates or the vacancy forecast. The 5% vulnerability figure, in particular, is a modeling assumption, not a measured outcome, and reasonable economists disagree sharply on how many roles AI will ultimately displace. The historical internet and smartphone analogies are also contested: those technologies created new categories of work; whether AI, which can itself perform cognitive tasks, follows the same pattern is the core question the industry has not resolved.
What to watch: whether AI-sector leasing activity in major markets, particularly in cities with established tech clusters, actually tightens prime availability in the next 12 to 18 months. That would be the earliest real-world test of CBRE's thesis, and it would show up in quarterly leasing data well before the 2031 vacancy target comes into view.