Forrester warns of shadow AI surge across Asia Pacific

Forrester has published its 2027 Asia Pacific predictions on the spread of AI in organisations across the region, with a focus on governance, control and accountability as use widens.

Among its main forecasts is that one in three large Asia Pacific enterprises will discover at least 10 employee-built shadow AI agents for every officially sanctioned agent. The firm expects employee-created agents, embedded software features and locally chosen AI platforms to spread faster than internal controls.

That is expected to create unmanaged risk, duplicated tools and undocumented dependencies inside large organisations. The report presents this as a shift from rapid AI deployment towards tighter oversight of how systems are approved, monitored and used.

Another prediction centres on model choice in regulated sectors. Forrester expects one in five large regulated enterprises in Asia Pacific to approve Chinese open-weight AI models, as companies seek greater control over model behaviour and where AI workloads run.

The forecast highlights a tension between operational sovereignty and geopolitical concern. Some organisations may accept self-hosted Chinese models because access to model weights and biases gives them more visibility, even as questions remain over provenance.

A separate forecast covers personal AI tools entering the workplace. Forrester predicts an employee's personal AI agent will trigger an incident at a Fortune 500 company, reflecting concerns that existing security systems can verify access rights but not always the intent behind automated actions.

As consumer AI products become more common, the boundary between personal and work use is likely to narrow. Employees may use personal agents to access corporate devices, credentials and data for work-related tasks, creating the risk that even a benign request could expose sensitive information or misuse access.

Human oversight

The predictions also extend to customer-facing businesses. Forrester expects 10 of the region's largest consumer-facing brands to try to stand apart through what it calls human assurance, particularly in sectors where trust carries more weight.

In practice, that means pairing AI-led interactions with access to human specialists for decisions involving financial risk, emotional vulnerability or complexity. The report points to financial services and healthcare as industries where human review or intervention is likely to remain central.

The broader argument running through the research is that AI in Asia Pacific is moving beyond the experimental phase and becoming more embedded in daily operations. As that happens, executives face pressure to contain costs and reduce risk without slowing internal adoption.

Forrester places particular emphasis on the diversity of the Asia Pacific market. Regulatory requirements, digital maturity and local expectations vary widely across the region, making a single approach to AI governance difficult for multinational companies and regional groups alike.

That diversity also shapes the sovereignty debate. In some markets, data residency rules remain the main concern. In others, attention is shifting to who controls the model itself, how inference is handled and whether companies can inspect the underlying system closely enough to satisfy internal risk teams.

The report forms part of Forrester's wider 2027 predictions research across technology, security, business buying, marketing and customer experience. Across those areas, the firm argues that AI will reshape spending priorities, infrastructure choices and the way companies interact with customers and employees.

Frederic Giron, Vice President and Senior Research Director at Forrester, said the region's mix of markets would make governance decisions more complex for leaders.

"In 2027, the challenge for Asia Pacific organizations will be to establish AI governance protocols in a region with distinctly diverse markets, regulatory requirements, and levels of AI readiness," said Frederic Giron, vice president and senior research director at Forrester. "Organisations that adapt their governance to local sovereignty requirements while investing in building long-term trust and resilience with their customers will be better positioned to capture AI's value."