AI is altering the way flexible workspaces handle bookings, property decisions, occupancy analysis and member support. However, the industry has not advanced uniformly. While a number of operators are reaping the rewards of connected digital systems, some are dabbling in AI with no dependable data to back it up or any clear return on investment.
What separates the field is the difference between AI’s potential, what operators are actually using, and what has delivered a measurable improvement. By and large, AI is not there to supplant the coworking staff. It is more valuable for putting information to better use, handling repetitive tasks, and meeting changing demand.
The Good - Smarter Operations and Better Decisions
Making commercial sense of workspace data
From room and membership bookings to billing, access, and customer calls, operators collect a lot of data. With AI analytics, operators can combine these figures to spot patterns that are hard to see manually. An operator might pinpoint which meeting rooms are not being used, predict when things will get busy or find out why some packages see more renewals than others. Such knowledge can guide pricing and staffing or a decision to put in phone booths or private offices where space is going to waste.
JLL’s 2026 Global Occupancy Planning Benchmark Report puts a focus on the need for accurate space data. Its research shows that, of the organisations it surveyed, only 8% have moved past the pilot stage in occupancy planning to something more active. A case in point: adoption is still in its early stages, and without good data, there is little to show for it.
Bookings on autopilot
An AI assistant can field member questions, lay out prices, and even make a reservation if tied to the booking software via an API. But do not mistake this for the whole story. The nitty-gritty of booking, payment and access control is usually handled by conventional software, not AI. That is a distinction worth noting in terms of cost and outcome.
Take Orbit Spaces in England. Since taking on the platform from its workspace management provider, the company has seen meeting-room revenue go up twofold in half a year and grow from one site to five in 18 months.
The system automated scheduling and payments that once had to be done manually. These are solid numbers from a well-run digital infrastructure, but they are hardly evidence that generative AI by itself is responsible for the gain. They are more of an illustration of how connected systems set the table for the more advanced stuff.
Building management
Facilities teams can use AI to look at building-management data for odd equipment behaviour or to put a figure on energy needs. Occupancy data can inform heating, lighting, and cleaning schedules. But it is a technical matter. To make a system useful, you need reliable sensors, historical readings, and integration with the BMS. And an operator has to know the difference between a room on the books and one that is actually in use.
CBRE’s March 2026 report, Data at the Heart of Hybrid, makes the point that workplace AI is only as good as the data underpinning it. There is opportunity here, but before talking of savings, operators should be measuring their operating costs and maintenance downtime.
The Bad - Costs, Data Problems and Uncertain Returns
Bringing disconnected systems together
It is common for operators to have distinct platforms for CRM, bookings, billing, facilities and door access. Adding a chatbot alone won't let AI bring these disparate systems into order. Proper integration requires the right APIs, authentication, role-based permissions, error handling and monitoring, and accurate records. A membership update that is late, for example, might cause a booking assistant to show false availability or an access system to work with old permissions.
Then there are the ongoing expenses: software, implementation, training staff, security and maintenance. Smaller operations should assess whether the outlay is justified by hard numbers on staff time, utilisation, member retention, or conversion.
JLL’s 2026 occupancy research gives some weight to this caution; most of the organisations they spoke to have yet to put AI into their occupancy plans. Buying tools before you have data quality and processes in order risks costly experiments instead of genuine operational gains.
The Ugly - Privacy, Unreliable Outputs and Lost Hospitality
When AI is at fault
Generative AI can be persuasive even when it is wrong. Within a coworking space, that means it may present incorrect pricing, get the cancellation terms or access details wrong. Operators must ensure their assistants work from approved, up-to-date data, test their output, and record any deficiencies. Sensitive matters should be in the hands of staff. For payments, contract changes, or access grants, the AI ought not to be trusted without question; its work requires explicit validation.
Then there is the matter of privacy. Booking histories, access logs and occupancy sensors can be revealing of an individual’s behaviour. Data must be minimised, retention kept within bounds, and vendor safeguards put in place. For those in the UK, the Data Protection Act 2018 and the UK GDPR are also considerations.
Automation in place of community
A coworking space is more than a set of desks. Community managers make introductions, put on events and solve problems so that members have reason to come back. While AI can free staff from some repetitive admin, overdo it and the interaction becomes impersonal. Algorithmic recommendations can also pick up bias in past data, as with lead-scoring that privileges profiles already on the books. The way forward is to let the machines handle the routine and leave relationships and judgement to people.
What Comes Next?
In its April 2026 report, The Flexible Office Space Imperative, JLL points to AI-induced uncertainty over workforce and organisational size as something that could fuel demand for flexible offices. Companies with uncertain headcount forecasts may prefer shorter-term commitments and a flexible workspace portfolio. But that is market logic, not evidence that AI has put a dent in coworking demand. The impact will depend on corporate real estate choices and investment.
Operators should zero in on a problem, set a baseline and see what happens. Put a number on the hours spent on booking administration, room use, energy, enquiry conversion and how satisfied members are. Only expand on the strength of the evidence.
AI does have a part to play in coworking, though not an even one. On the plus side, there is better information and efficiency. On the downside, there are the cost of the technology and its upkeep. And then there's the ugly side of trusting a system you can't rely on, or giving up human contact for the sake of automation. Those who win will be the ones who put AI to work to improve the workspace experience, not just to say they are doing so.
