Ocean: The AI Email Security Startup Using ‘Agentic’ Models to Stop AI‑Generated Phishing

AI‑generated phishing is becoming one of the fastest‑growing cybersecurity threats, and a new startup believes traditional email defenses are not designed for it. Ocean, a cybersecurity company founded in 2023, has emerged from stealth with $28 million in funding to build an AI‑native platform that analyzes the intent behind emails rather than just scanning for known malicious patterns.

The company’s approach relies on what it calls an “agentic email security platform,” where networks of AI agents investigate each incoming message to determine whether it is trying to manipulate a recipient.

The startup behind the idea

Ocean was founded by Shay Shwartz, a cybersecurity researcher with an unconventional background. As a teenager, he reportedly experimented with hacking before being caught at age 16 and redirecting his skills toward defensive security work.

He later spent about a decade working in Israel’s elite defense and intelligence units, including roles connected to the Iron Dome missile defense system, before moving into the private sector and eventually launching Ocean in 2023.

Two years after starting the company, Ocean publicly launched with $28 million in total funding, including a $20 million Series A round led by Lightspeed Venture Partners with participation from Picture Capital, Cerca Partners, and several prominent angel investors.

What Ocean actually builds

Ocean focuses on one problem: email attacks generated or enhanced by artificial intelligence.

Traditional email security tools typically rely on techniques such as:

• Known malicious signatures
• Reputation databases
• Pattern or rule matching

Those approaches can work against repeated attack templates. But AI‑generated phishing emails can vary tone, wording, and structure almost infinitely, allowing them to bypass filters that depend on recognizable patterns.

Ocean’s platform instead tries to understand the meaning and objective of each email.

How Ocean’s “agentic” AI model analyzes emails

The company describes its system as an agentic network of AI agents coordinated by a central intelligence engine. These agents investigate emails before they reach a user’s inbox, collecting context and signals across multiple dimensions.

According to the company’s platform description and reporting on its launch, the process broadly works like this:

• Small language models (SLMs) and larger models help interpret the text of the email.
• Multiple specialized AI agents analyze different aspects of the message, such as language, context, and intent.
• The system gathers additional signals and context, enriching the analysis before deciding whether the message is malicious.

Rather than simply asking “does this email look like a known phishing template?”, the system tries to answer a deeper question: “What is this email trying to get the recipient to do?”

If the AI concludes the message is attempting manipulation—such as requesting credentials, transferring money, or triggering a fraudulent workflow—it can block or flag the message before it reaches the user.

Why AI‑driven phishing is harder to stop

The rise of generative AI has made social‑engineering attacks dramatically easier to produce and personalize.

Large language models allow attackers to generate messages that:

• Mimic internal corporate tone
• Reference specific projects or colleagues
• Avoid grammatical errors typical of older phishing attempts

Because of this variability, defenders increasingly argue that pattern‑matching filters alone are insufficient. Ocean’s pitch is that understanding intent and context—rather than static indicators—is necessary to detect modern attacks.

Early traction and adoption

Ocean launched publicly with $28 million in total funding, with its Series A led by Lightspeed Venture Partners.

According to company statements reported around its launch, the platform is already protecting hundreds of thousands of mailboxes while the company continues expanding its product and customer base.

More detailed metrics about total emails analyzed, number of customers, or long‑term adoption rates have not yet been widely disclosed.

Why investors are paying attention

The company’s funding round reflects a growing belief among security investors that AI will transform both sides of the cybersecurity battle.

Attackers now use AI to generate highly convincing phishing messages at scale, while defensive platforms like Ocean aim to counter them with AI agents capable of investigating every email automatically.

Whether agent‑based email security becomes the dominant approach remains to be seen, but the scale of Ocean’s launch funding suggests investors expect the arms race between AI‑powered attackers and AI‑powered defenders to intensify.