Wand Team
ZeroDrift extends the Wand AI governance layer with pre-send compliance enforcement for AI-generated communications in financial services.
ZeroDrift, The Enforcement Runtime for AI, today announced the general availability of Anchor 3.0, the first family of small language models built to enforce regulatory and company rules on AI-generated communications before they’re sent. The company also published the first benchmark measuring how well AI models enforce FINRA rules on business communications, built on human-written, attorney-labeled data independently produced by Surge AI, the data labeling company trusted by the world’s leading AI labs. Anchor 3.0 caught 95.5% of violations, more than any frontier model tested. It also outperformed GPT-5.6 Sol, the strongest frontier model on the test, on recall, precision and F1 — a measure that balances recall and precision — while running up to 34x faster and up to 12x cheaper. Developers can sign up and start using Anchor 3.0 today.
“ZeroDrift extends the Wand AI governance layer to incorporate external regulatory requirements, enabling our agents’ communications to be checked and corrected before they are sent,” said Cristian Felix, Chief AI Architect of Wand AI. “Strong governance is essential for deploying AI at scale, giving financial institutions the confidence and control they need to move from experimentation into production.”
Financial institutions are already moving AI from experimentation into production. Cambridge research found that 81% of surveyed financial services firms are adopting AI at some level, with traditional financial institutions reporting 45% agentic AI adoption. report notes that generative AI can implicate supervision, communications, record-keeping, and fair dealing obligations, and regulators have levied billions of dollars in fines for communications failures. Frontier models can catch many violations, but at 12 to 51 seconds per message for the strongest of them, they can only review after the fact. Anchor 3.0 runs the check on every message before it leaves, in about 1.5 seconds on ZeroDrift’s API and under 100 milliseconds self-hosted.
"Frontier models made it easy to build capable agents. The hard part is running them inside a regulated business, where every message has to follow the rules and the check has to happen every time, before anything goes out," said Kumesh Aroomoogan, founder and CEO of ZeroDrift. "A frontier model is too slow and too expensive to do that on every message. Anchor is built for exactly this job. Developers keep building with the models they choose, and Anchor makes sure what their agents say is compliant."
Anchor 3.0 is available in three sizes:
-
Anchor 3.0 Mini — 9B parameters (4B active), post-trained from Gemma E4B. The fastest and lowest-cost model in the family, built for the highest-volume AI traffic. Runs ZeroDrift’s pre-built rule packs and flags violations.
-
Anchor 3.0 — 9B parameters (4B active), post-trained from Gemma E4B. The flagship model and the one behind the headline results. Runs ZeroDrift’s full library of more than 200 pre-built rules across FINRA, SEC and other regulations, flags the exact lines that violate a rule and rewrites them so the message goes out compliant. Custom company policies are trained into the model through a LoRA adapter.
-
Anchor 3.0 Max — 27B parameters, post-trained from Qwen3.8-27B. The most capable model in the family, built to enforce a company’s own policies out of the box with no fine-tuning. Handles long-form content and document attachments with the largest context window in the family.
Developers can sign up at https://zerodrift.com/ and start enforcing AI communications immediately through the ZeroDrift Enforcement API. The full model-by-model FINRA benchmark results and methodology are available now.
About Wand AI
Wand AI is building the sovereign infrastructure for AI labor. Wand's operating system enables governments and enterprises to deploy, manage, govern and continuously evolve AI agents alongside humans as trusted members of the workforce, with identity, objectives, authority, security, governance and auditability built into the platform.
For nations, Wand provides a unified foundation through which ministries, public institutions, and critical national organizations can deploy AI labor while maintaining sovereign control over models, compute, data, policies and operations. Wand powers production-scale deployments with some of the world's most consequential institutions, including leading banks, asset management firms, hedge funds, consulting firms, and system integrators. Wand exists to power the biggest transition in human history: infinite labor—so that everything humanity can imagine, we can finally build.
Founded in 2023 and headquartered in Palo Alto, California, with offices in Palo Alto, New York and Abu Dhabi, Wand is backed by a tier-one research team and world-class investors, including Shasta Ventures, Fusion Fund, Thiel Capital, and Palo Alto Growth Capital. Learn more at wand.ai.
About Zerodrift
ZeroDrift is the company that pioneered enforcement runtime for AI communications. As AI begins to speak and act on behalf of organizations, ZeroDrift is the infrastructure that enforces the rules those organizations live by. The platform comes with every major regulation pre-loaded, and a company’s own policies trained on top, and it controls what AI agents say, see, hear and do before they communicate externally or take action, with a record a regulator can audit. ZeroDrift is starting with compliance on AI communications and expanding to every rule a company writes, across every AI surface. Founded in New York by AI and platform leaders from Google DeepMind, Microsoft AI and Goldman Sachs, ZeroDrift is backed by $10 million from a16z speedrun, PitchDrive Ventures, and other leading AI investors. Learn more at ZeroDrift.com.
