REPORT
The SpyCloud 2026 Forecast Report:
Lessons from 2026, Predictions for 2027
Eight SpyCloud experts and 50 enterprise security pros weigh in on where identity threats are heading next, and how last year’s predictions held up.
While we take a look back at 2026, we’re also shining light on the threats we see emerging – or continuing – into 2027. The SpyCloud 2026 Forecast Report draws on intelligence from SpyCloud’s in-house experts, grades last year’s predictions, and puts fresh calls for 2027 on the table – ranging from the near-certainities to some hotter takes.
Six of our ten 2026 predictions held up, and the other four continue to develop (talk about being ahead of our time). This year’s forecast features eight SpyCloud experts across threat research, investigations, product, and data science, while incorporating survey data from 50 security professionals at large enterprises across a range of industries.
The result is a view of 2027 that pairs what attackers are doing with what defenders are actually facing – and leaves room for what you can do to prepare.
Inside this report you’ll find:
Defenses under pressure
How attackers keep finding ways around the controls organizations rely on, including MFA, session protections and device trust
Non-human identities and AI agents
The growing exposure of the keys, tokens, and service accounts that power automation and AI tools, and why they’re becoming a prime target
Access you already trust
How vendors, integrations and developer tooling keep opening doors to the biggest compromises
Insider threats, redefined
Why the line between employee, contractor, and outsider keeps blurring, and what security teams aren’t yet planning for
Synthetic identity and fraud
How stolen data and AI are making fabricated identities cheaper to build and harder to spot
Shifting threat actor communities
How groups, brands, and chat platforms keep changing, plus what that means for tracking the people behind the attacks
AI's real role in cybercrime
Where AI is speeding up familiar attacks, where the headlines are outpacing the evidence, and how our experts weigh the difference
Key Takeaways:
-
MFA is routinely bypassed:
48% of surveyed organizations experienced account takeovers despite having MFA enabled, primarily through session and device hijacking. -
AI Tokens are among the new targets:
Infostealers like WardenStealer are pivoting to collect API keys and tokens used by AI agents and coding tools. -
Trusted access risks persist:
68% of enterprises were affected by vendor or SaaS breaches, highlighting that "trusted" integrations are now a primary attack vector. -
Synthetic identities evolve:
Fraudsters are combining real stolen data with AI-generated media to create highly convincing synthetic identities. -
Phishing AI agents:
As organizations adopt AI agents with internal system access, attackers are beginning to phish these non-human identities (NHIs).
About our research
This forecast draws on what SpyCloud sees in the criminal underground, our SpyCloud Labs research and monthly cybercrime updates, and over a decade of disrupting cybercrime. SpyCloud experts from across the business share what they’re seeing and what security teams should have on their radar in 2027.
Also included is the analysis of a focused survey run in October 2026 – specifically, security professionals at large enterprises across industries. Their answers show what teams dealt with in 2026 and what’s on their minds heading into 2027 – so you can see where the experts and the field agree, and where they don’t.
SpyCloud’s identity threat protection platform helps organizations find and remediate exposed identities before criminals can use them. Keep reading about the current landscape in the 2026 Identity Threat Report.
Check Your Exposure
See your exposure details and get one more step ahead of attackers today.