AI agents and machine identities will widen 2026 attack surfaces

At a glance

What this is: This is Expel's 2026 security forecast, and its central finding is that AI, agentic tooling, and machine identities will intensify existing identity and access control weaknesses rather than replace them.

Why it matters: It matters because IAM, PAM, and NHI teams will be asked to govern more credentials, more delegated access, and more opaque automation without assuming current review and monitoring models are enough.

👉 Read Expel's 2026 cybersecurity predictions on AI, identity, and attacker behaviour

Context

The security gap here is not the arrival of AI alone. It is the combination of AI adoption, interoperability, and long-lived credentials that gives attackers more opportunities to exploit access paths that were already difficult to govern. In identity terms, that means machine identities, API keys, and agent delegation are becoming a larger share of the control problem, not a separate one.

Expel's forecast points to a familiar pattern in a new wrapper: defenders are still struggling with basic access hygiene while attackers iterate quickly against exposed identities, misconfigurations, and third-party trust. For IAM and NHI practitioners, this is a governance issue as much as a detection issue, because visibility and lifecycle control lag behind how quickly AI-enabled systems are being deployed.

Key questions

Q: How should security teams govern AI agents that can access enterprise systems?

A: Security teams should govern AI agents as non-human identities with explicit ownership, scoped privileges, and continuous monitoring. The control set should include inventory, task-bound credentials, audit trails, and revocation paths. If an agent can call tools or touch production systems, it belongs in the same governance model as service accounts and other machine identities.

Q: Why do machine identities create more risk than human identities in some environments?

A: Machine identities are often numerous, long-lived, and embedded in code or infrastructure. They are harder to review manually, easier to overlook during offboarding, and more likely to carry excessive privilege. That combination increases blast radius when a secret or token is exposed.

Q: How should security teams handle exposed API keys and service credentials?

A: Treat exposed credentials as active identities, not as misplaced text. Revoke or rotate the secret immediately, identify every service it can reach, and then confirm which workflows break. A credential leak is only resolved when the identity is no longer valid anywhere it was trusted, including third-party integrations and automation paths.

Q: How do IAM and PAM teams apply governance to agentic AI testing platforms?

A: Treat the agent as a delegated actor with bounded authority. That means scope limits, revocation conditions, approval gates, and traceability should be designed like privileged access controls, not left as product settings. If an agent can act on behalf of the organisation, the governance model should resemble controlled delegated access.

Technical breakdown

Why machine identity sprawl is the real 2026 control problem

Machine identity sprawl occurs when service accounts, API keys, tokens, and certificates multiply across tools, pipelines, and AI workflows faster than governance can classify them. Expel's forecast reflects a broader reality: AI-driven interoperability creates more integration points, and each one introduces credentials that are often long-lived and poorly inventoried. The technical issue is not just volume, but boundary loss, because identities begin to cross application, cloud, and vendor lines without clear ownership or revocation discipline.

Practical implication: inventory machine identities by owner, purpose, and expiry, then remove any credential that cannot be tied to a lifecycle control.

How agentic tooling changes access control assumptions

Agentic tooling changes access patterns because the system making decisions is not a human user, but software that can call tools, chain actions, and retain context across sessions. That makes traditional approval and review models less reliable, especially when agents are granted broad scopes to preserve usability. The result is a growing mismatch between intended privilege and actual runtime behaviour, which is why agent governance must consider delegation boundaries, scoped credentials, and observable action trails.

Practical implication: bind each agent to task-scoped permissions and require auditable tool-use logs before allowing broader delegation.

Why attacker tempo keeps beating patch and detection cycles

Expel's contributors describe a familiar asymmetry: attackers do not need novel techniques when weak identity controls, exposed secrets, and delayed patching still work. In practice, this means compromise can happen minutes after exposure, while defenders often rely on slower review and remediation cycles. For identity teams, the challenge is not only blocking initial access but reducing the time a usable secret remains valid after discovery or leakage.

Practical implication: shorten credential lifetime, automate revocation, and treat exposed secrets as immediate containment events rather than queue items.

NHI Mgmt Group analysis

Machine identity governance will become a board-level control issue, not a niche IAM concern. Expel's forecast is consistent with what we see across NHI programmes: the number of non-human credentials grows whenever interoperability and automation are rewarded faster than governance is redesigned. The new concept here is identity delegation drift: permissions granted for one workflow gradually become standing access across multiple workflows. That drift is what turns AI adoption into an access problem, and practitioners need lifecycle controls that match the pace of deployment.

Agentic AI exposes a trust gap between decision-making and accountability. Once software can choose tools and timing, the question is no longer whether an identity exists, but whether its actions are bounded, attributable, and revocable. That is where IAM, PAM, and NHI controls converge: ownership, least privilege, and runtime logging all need to hold even when the actor is non-human. Practitioners should treat agent governance as an extension of privileged access governance, not a separate innovation track.

Attackers will keep winning with simple identity abuse because defenders still overestimate the value of complexity. Expel's contributors are right to downplay the idea that AI automatically changes the attacker playbook. The main shift is scale and speed, not novelty. That means credential exposure, over-privilege, and weak third-party visibility remain the most dependable failure modes, and teams that ignore those basics will keep losing ground.

AI strategy is increasingly inseparable from identity strategy. Expel's outlook shows that the practical risk is not just model misuse or application abuse, but the identities that let AI systems touch data, APIs, and production workflows. The category is maturing around governance questions that IAM leaders already know well: who owns the identity, what can it do, when does it expire, and how is misuse detected. Practitioners should align AI rollout reviews with identity governance reviews from the start.

Standing credentials remain the weak point that turns experimentation into exposure. The article repeatedly points to long-lived access, exposed keys, and uncontrolled third-party connections as the conditions attackers exploit. In NHI terms, that is the persistence of unmanaged trust. Teams need to recognise that AI expansion magnifies a pre-existing governance debt, and the right response is to eliminate standing privilege where it is not strictly necessary.

What this signals

Expanding AI use will force identity teams to treat machine identities as first-class citizens in access governance, especially where API keys and delegated access already outnumber human users. The programmes that do best will be the ones that can answer a simple question quickly: who owns this non-human credential, and when does it expire?

Identity delegation drift: this is the pattern where temporary automation permissions become permanent operating assumptions. It will matter most in environments that mix SaaS integrations, internal agents, and third-party workflows, because the attack surface expands faster than review cadences can keep up.

For practitioners, the near-term signal is clear: security posture will be judged less by whether AI is deployed and more by whether access to AI systems is governed through lifecycle controls, revocation, and monitoring. The control baseline is shifting toward Top 10 NHI Issues and the lifecycle discipline captured in the Ultimate Guide to NHIs , Key Challenges and Risks.

For practitioners

• Map all machine identities to owners and expiry rules Create an inventory of service accounts, API keys, tokens, and certificates used by AI tooling, then assign a business owner, technical owner, and expiration policy to each one.
• Restrict agent permissions to task-scoped operations Require agents to use narrowly scoped credentials tied to a specific workflow, with separate approvals for data access, tool execution, and production changes.
• Automate revocation for exposed or abandoned secrets Treat leaked keys, unused tokens, and orphaned certificates as immediate containment items, with automated rotation and revocation triggered by exposure signals.
• Review third-party and OAuth-connected access paths Reassess connected apps, delegated access, and vendor integrations to confirm that each trust relationship has a current owner, purpose, and offboarding path. This is where hidden NHI exposure often accumulates.

Key takeaways

• Expel's 2026 outlook reinforces that AI expansion is primarily an identity governance problem when machine credentials, agent delegation, and third-party trust multiply at speed.
• The most durable risks remain familiar: standing access, exposed secrets, and weak visibility into non-human identities continue to give attackers the easiest path.
• Teams that align AI rollout decisions with NHI lifecycle controls, revocation, and privileged access review will be better positioned than those treating AI as a separate security domain.

Key terms

• Machine Identity: The digital identity of a machine, device, or workload — such as a server, container, or VM — used to authenticate it within a network. Sometimes used interchangeably with NHI, though NHI is the broader category.
• Agentic AI: Autonomous AI systems capable of planning, deciding, and taking actions — including calling APIs, writing code, and orchestrating other agents — with minimal human oversight. Agentic AI introduces new NHI risks as agents must authenticate to external services.
• Identity Delegation Drift: Identity delegation drift is the gradual expansion of permissions originally granted for a narrow workflow into broader, persistent access. It often appears in automation and AI tooling where convenience overrides review, leaving organisations with standing trust they no longer understand or need.
• Standing Privilege: Standing privilege is access that remains active even when no immediate task requires it. For NHI programmes, it is a common failure mode because long-lived credentials and persistent roles create unnecessary exposure. Reducing standing privilege usually means tighter expiry, on-demand access, and clearer review of who or what still needs access.

What's in the full article

Expel's full forecast covers the operational detail this post intentionally leaves for the source:

• Direct quotes from executives and field leaders on AI, geopolitics, and regulation
• The broader 2026 prediction set across federal policy, attacker behaviour, and developer targeting
• Practitioner context on how Expel's team is weighting identity, patching, and AI security in its outlook
• The source article's full discussion of where AI changes the security workload versus where it does not

👉 Expel's full forecast adds the leadership commentary and broader 2026 threat outlook behind these findings.

Deepen your knowledge

The NHI Foundation Level course, the industry's only accredited NHI security programme, covers NHI governance, machine identity security, secrets management, and agentic AI identity. It is designed for practitioners building control models that can keep pace with automation and delegated access.