Harbor Raised $10M to Insure AI Agents in Its First Year. Here Are the Deck and the Memo

Theme 1: AI agents have created an insurance gap that traditional carriers can't price

Insurers lack the loss data to underwrite agent risk

The core problem is actuarial: insurance prices risk from claims history, and agents don't have one yet.

"The company's insurer can't price it, because insurers price risk from years of claims, and agents have almost none."

Carriers are actively retreating from AI exposure

Rather than learning to price the risk, many insurers are excluding it from policies companies already hold, which widens the gap.

"Since January, US carriers have been able to attach generative-AI exclusions to general liability policies, and many have started carving AI out of the cover companies already pay for."

Liability for agent errors is unresolved

The article frames the market through a scenario where accountability is unclear and errors compound at machine speed.

"An AI agent at a bank approves the wrong payment, a few thousand times, before anyone notices. Who pays?" "Today the honest answer is that nobody knows."

Theme 2: The winning model is measure, prevent, then insure (not insurance alone)

Harbor bundles risk measurement, enforcement, and coverage

Harbor's approach reduces the insurable risk before underwriting it, which helps solve the missing-loss-data problem.

"Harbor measures the risk of every agent a company runs, blocks the decisions that break the company's own rules, and insures what's left."

A live risk score is positioned over questionnaires and audits

The paywalled memo's pitch is that continuous measurement beats point-in-time assessment.

"The Memo: the one-pager that explains why a live risk score beats a questionnaire or a one-off audit"

A competitor is converging on the same insight from the certification side

AIUC ties insurance to audit, suggesting the category is validating around "verify, then insure."

"AIUC, which certifies agents and ties insurance to the audit, raised a $40M Series A led by Ribbit in September."

Theme 3: The AI-agent risk market formed remarkably fast, pulled by regulation

Regulatory deadlines create urgency

The ECB has set a concrete forcing function for banks, a key buyer segment.

"The ECB has asked every bank it supervises for a plan against AI-enabled threats, due October 31."

The market went from nonexistent to funded in under three years

The article's own timeline chart caption frames the speed of formation.

"In 32 months the AI-agent risk market went from a chatbot court ruling to two funded startups and a regulator deadline." "...in a market that barely existed three years ago"

Theme 4: Timing plus preparation can overcome a "wrong-profile" team

Market timing did half the work

"The timing did half the work."

Pre-empting investor objections did the rest

The author's central argument is that the fundraising playbook, not the credentials, is the replicable part.

"The other half was preparation. Every reason an investor could give for passing on this team had a named answer before the first meeting, and you can see each one in the deck."

2. Contrarian Perspectives

A team with no insurance background can raise in a heavily regulated category

Conventional wisdom says regulated verticals require domain-veteran founders. Harbor's raise cuts against that, though the author discloses a conflict of interest.

"None of them had run an insurance business. On paper, that profile is a hard sell in a category this regulated. They raised $10M anyway, from three lead investors..."

Supporting evidence: the founders' backgrounds ("trading desk, consulting, robotics and AI safety research"), only one prior founder, and a round with a deep syndicate (Base10, Abstract, 8VC, NFX, Restive, Innovation Endeavors, Kibo Ventures).

"No data" can be treated as the opportunity rather than the blocker

Most would see zero loss history as a reason insurance can't exist yet. The framing here is that an empty actuarial record is a gap a new entrant can fill by generating its own risk data.

"Four founders who met at Stanford, three co-leads and a market with zero loss data. The moves you can copy"

Supporting evidence: the strategy of measuring every agent's risk and blocking rule-breaking decisions before insuring "what's left" converts an unpriceable exposure into a managed one.

3. Companies Identified

Harbor

• Description: AI-agent risk startup that measures agent risk, blocks rule-violating decisions, and insures the remainder. Came out of stealth with a $10M seed.
• Why mentioned: Subject of the article and a case study in raising on a "wrong-profile" team; the author is an investor.
• Quotes: "Harbor measures the risk of every agent a company runs, blocks the decisions that break the company's own rules, and insures what's left." / "I'm an investor in Harbor, so read this with that in mind."

AIUC

• Description: Company that certifies AI agents and ties insurance to the audit.
• Why mentioned: Direct competitor and market validation; raised a $40M Series A.
• Quotes: "AIUC, which certifies agents and ties insurance to the audit, raised a $40M Series A led by Ribbit in September."

Base10

• Description: Venture investor, co-lead of Harbor's seed.
• Why mentioned: Co-lead investor.
• Quotes: "Base10, Abstract and 8VC co-led"

Abstract

• Description: Venture investor, co-lead of Harbor's seed.
• Why mentioned: Co-lead investor.
• Quotes: "Base10, Abstract and 8VC co-led"

8VC

• Description: Venture investor, co-lead of Harbor's seed.
• Why mentioned: Co-lead investor.
• Quotes: "Base10, Abstract and 8VC co-led"

NFX, Restive, Innovation Endeavors, Kibo Ventures

• Description: Venture investors participating in the round.
• Why mentioned: Part of the seed syndicate.
• Quotes: "with NFX, Restive, Innovation Endeavors, Kibo Ventures and others in the round."

Ribbit

• Description: Investor that led AIUC's $40M Series A.
• Why mentioned: Signals institutional appetite for the category.
• Quotes: "raised a $40M Series A led by Ribbit in September."

ECB (European Central Bank)

• Description: Regulator supervising European banks.
• Why mentioned: Its AI-threat planning requirement is a demand catalyst.
• Quotes: "The ECB has asked every bank it supervises for a plan against AI-enabled threats, due October 31."

4. People Identified

Ruben Dominguez

• Description: Author of The VC Corner and investor in Harbor.
• Why mentioned: Writer of the piece; discloses his stake.
• Quotes: "I'm an investor in Harbor, so read this with that in mind. I'm writing about it because of how the round came together, which is the part most founders can copy."

Harbor's four co-founders (unnamed)

• Description: Stanford-connected team with backgrounds in trading, consulting, robotics, and AI safety research; one previously founded a company.
• Why mentioned: Case study of a team that raised despite lacking insurance experience.
• Quotes: "Its four co-founders met at Stanford, and their CVs read trading desk, consulting, robotics and AI safety research. One of them had founded a company before, and none of them had run an insurance business."

5. Operating Insights

Build an objection map before your first investor meeting

List every reason an investor could pass on you and attach a named answer (a person, partner, or mechanism) to each. For this team, the gap was domain experience, and it was closed with specific mitigations rather than hand-waving.

"Every reason an investor could give for passing on this team had a named answer before the first meeting, and you can see each one in the deck." "The Five-Gap Map: every objection an investor would raise about a first-year team, and the person, partner or mechanism Harbor put against each"

Stress-test your own deck before sending it

Run the deck through an adversarial gap audit first, so weaknesses are found internally rather than in partner meetings.

"The Gap Audit Prompt: a Claude prompt that runs your own deck through the same test before you send it to a single fund"

Anchor the why-now to a dated external forcing function

Tie urgency to regulatory deadlines, competitor funding, and policy changes rather than abstract trend claims. Harbor's deck leads with a why-now slide and a concrete six-month plan.

"The $10M Seed Deck: the full deck Harbor used to raise, from the why-now slide to the six-month pla[n]"

6. Overlooked Insights

The generative-AI exclusion shift may be forcing companies into the market involuntarily

The mention that carriers can now exclude gen-AI from general liability policies implies enterprises may be losing coverage they assumed they had, creating a latent buyer base that doesn't yet know it needs new coverage.

"Since January, US carriers have been able to attach generative-AI exclusions to general liability policies, and many have started carving AI out of the cover companies already pay for."

Three co-leads is an unusual structure that may itself be a de-risking signal

Most seed rounds have one lead. A three-way co-lead in a novel, regulated category suggests syndicate-level validation that reduces perceived risk for each participant, an angle the article touches on only in passing.

"They raised $10M anyway, from three lead investors, in a market that barely existed three years ago"