Microsoft's 2026 Digital Defense Report calls for stronger governance of agent identities and sensitive data as AI expands both the targets available to attackers and their capabilities. Its account distinguishes attacks against AI systems from attacks on conventional systems that use AI to increase speed or scale.

The report identifies risks around manipulated instructions, exposed data, compromised privileges, excessive agent actions and operational integrity. Microsoft links these to controls including scoped credentials, limits on tool use and changes, and records that can withstand tampering. These are the company's recommendations, rather than proof that any particular product prevents all such attacks.

For financial institutions, the practical question is whether an agent's authority is visible and enforceable throughout a workflow. A user may be entitled to view customer records while still requiring separate approval to amend them or initiate a payment.

Security reviews should therefore follow a complete task, including tool calls and handoffs between agents. Testing only the model's final answer can miss the point at which an instruction becomes a consequential action. The report offers a useful framework for asking where that authority is granted, checked and withdrawn.

Execution Level Governance- What audit-ready agent governance actually looks like
David Girvin, founder and CEO of Assury argues that model-in-the-loop review, AI governing AI, is fundamentally unreliable for regulated environments: even the best-performing models miss a meaningful share of violations, the reviewing model is typically provided by the same vendor being reviewed, and prompt injection or context poisoning can compromise both the acting agent and its supposed overseer simultaneously. He makes the case for deterministic, architecturally enforced controls instead, walking through Assury’s approach of autonomy zones, session risk accumulation, and credential starvation, which lets a compromised agent be cut off from its tools instantly rather than relying on time-boxed access. The conversation touches on why David is sceptical of just-in-time credentialing as a solution for agent security more broadly, since agent sessions don’t run on predictable human timescales, along with the current gap between how identity and security vendors are pitching agent protection and what he sees happening at the execution layer in practice. He also discusses the compliance and audit implications of probabilistic decision-making, arguing that regulated industries will increasingly need tamper-evident, hash-chained audit trails that can withstand scrutiny from auditors and regulators who are only beginning to understand agentic risk, and reflects on a named frontier lab’s own published framework as an example of the gap between research and practitioner reality. Elsewhere, David reflects candidly on building a bootstrapped security company in an increasingly crowded market, why he turned down aggressive VC funding to stay in control of the product, and what a credible third-party assessment of his own gateway would need to look like given that Assury sits directly in the execution path for every customer’s agents.
BrightTALK
Share this post
The link has been copied!