Trust3 AI heads to The AI Conference with agent security pitch

Trust3 AI is showcasing its Agent DOS control plane at The AI Conference in San Francisco on Sept. 30 and Oct. 1 as enterprises face mounting pressure to secure AI agents before moving them into production. The company is betting that visibility, access control and data-layer enforcement will determine whether AI projects deliver ROI or stall in security review.

Why it matters: - Enterprises are moving from AI experimentation to AI agents that can read records, take actions, call tools and hand work to other agents. - Trust3 AI is pitching agent security as the gatekeeper for production AI, arguing that ROI stalls when security teams cannot approve what they cannot observe. - The company says its control plane is designed to help enterprises move agents from pilot to production faster while reducing audit friction.

What happened: - Trust3 AI is at The AI Conference at Pier 48 in San Francisco, Booth 6 in the Innovation Hub, on Sept. 30 and Oct. 1. - The company is demonstrating Agent DOS, its control plane for discovering, observing and securing agents and attack surfaces. - Trust3 AI is using the event to make the case that AI agent security, not model capability alone, is now the core enterprise question. - The company also points to California's new executive order and recent AI security incidents as signs that agent safety has become a public and board-level issue.

The details: - Trust3 AI says Agent DOS helps teams discover and assess every agent, including agents built and connected to data without a security review. - Agent DOS also observes each decision from prompt to tool call to output destination, so reviewers can trace what an agent did and why it was allowed. - The platform uses purpose-based access control, or PBAC, evaluated per request, so an agent gets only the access its task requires. - Trust3 AI says every agent can also have a kill switch. - The company says enforcement happens natively at the data source on Snowflake, Databricks, BigQuery, Iceberg and more than 50 other platforms. - Trust3 AI says it carries the originating user's identity and purpose through each hop. - The company says its customers include a Fortune 500 financial firm that discovered more than 300 agents during an audit and became audit-ready without pausing operations. - Trust3 AI says a Fortune 100 fintech replaced over-privileged AWS IAM roles with attribute-based access. - Trust3 AI says a top-5 U.S. healthcare retailer governs thousands of BigQuery tables across hundreds of GCP projects without a ticketing queue. - The company says it helps teams move from proof of concept to production 10x faster and cuts audit preparation time by 84%. - Trust3 AI says it is a Snowflake Startup Accelerator member, Databricks Built On Partner, AWS Qualified Software, Microsoft Partner, Google Cloud Select Technology Partner and NVIDIA Inception member. - The company says investors include Point72 Ventures, Sapphire Ventures, Insight Partners, Battery and Accel. - Trust3 AI provided a demo link at More information. - Trust3 AI also directs readers to Learn more.

Between the lines: - The message is aimed at security, risk and governance teams, not just AI builders. - The company is framing agent security as a condition for scale, not a compliance add-on. - Recent incidents cited in the release show a common failure mode: systems controlled access, but did not govern what agents did after access was granted. - Trust3 AI is positioning itself against that gap by emphasizing identity, declared purpose and policy enforcement at the data layer. - The company is also leaning on market warnings, including Gartner's projection that more than 40% of agentic AI projects will be canceled by the end of 2027.

What's next: - Trust3 AI plans live demonstrations at Booth 6 throughout the conference on Sept. 30 and Oct. 1. - The company is steering attendees to ask how many agents are running, what each agent is allowed to do, where enforcement happens and what stops an agent if it exceeds scope. - Trust3 AI's longer-term bet is that enterprises will adopt agent controls now to avoid access-control failures later, especially as AI attacks mature. - Gartner expects most successful attacks on AI agents by 2029 to come through access control weaknesses.