Ai2 launches Shippy Artificial Intelligence agent for ocean monitoring

Seattle’s Allen Institute for Artificial Intelligence has introduced Shippy, a free agent for maritime analysts using its Skylight ocean-monitoring platform. The tool answers plain-language questions about vessel activity while linking responses to underlying records.

Skylight, the free ocean-monitoring platform built by Seattle’s Allen Institute for Artificial Intelligence, launched Shippy, an Artificial Intelligence agent designed to answer plain-language questions about activity across the world’s oceans. The tool draws on Skylight’s live vessel-tracking and satellite data to help maritime analysts investigate issues such as illegal fishing and vessels that have gone dark. Each answer links back to the underlying records so analysts can verify and reproduce the results.

Skylight is one of a group of environmental projects that moved in 2021 to Allen Institute for Artificial Intelligence from Vulcan Inc., now known as Vale Group, the late Microsoft co-founder Paul Allen’s holding company. Researchers at Skylight have spent years building tools to spot illegal, unreported and unregulated fishing, which it says accounts for billions of dollars in losses each year and hits developing countries that depend on their fisheries the hardest. The platform is free, and Skylight says it is used by more than 300 organizations across about 70 countries.

The system combines free satellite data with commercial imagery and vessel-tracking feeds to flag suspicious behavior, including a ship going dark or two vessels meeting at sea to transfer catch. Allen Institute for Artificial Intelligence has open-sourced the computer-vision models behind the project. Like Skylight itself, Shippy will be free to governments, regional fisheries bodies and qualifying nonprofits.

Access is currently limited to a small group of agencies and partners, with Skylight planning to expand availability as the tool is updated and improved. Built-in limits are intended to keep Shippy accountable: it stays focused on maritime questions, presents facts without making legal judgments, declines defense-related requests, and stops rather than guess when its data cannot support an answer. Operational decisions, such as where to send a patrol, remain with human users.

52

Impact Score

Banking CISOs face artificial intelligence governance gap

Banking security leaders are moving quickly to formalize Artificial Intelligence oversight as business deployments and examiner scrutiny increase. Microsoft Copilot, agentic platforms, and third-party tools are turning governance gaps into operational risk.

Apple delays Siri Artificial Intelligence in EU amid DMA dispute

Apple says its redesigned Siri Artificial Intelligence will not launch on iPhones or iPads in the European Union under upcoming operating system releases. The company blames an unresolved dispute with regulators over DMA requirements and user privacy protections.

UK unveils £1.1 billion Artificial Intelligence hardware plan

The UK government is backing chip firms, computing infrastructure and skills with a £1.1 billion Artificial Intelligence Hardware Plan. The package includes a £750 million national Artificial Intelligence supercomputer and startup support tied to next-generation chips.

Kirkland brings Artificial Intelligence ambitions to Palantir stage

Kirkland & Ellis used a Palantir conference to showcase a fund formation platform designed to automate major parts of private funds work. The presentation underscored Big Law’s accelerating Artificial Intelligence push while leaving pricing and business model questions unresolved.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.