Haven Safety AI Launches AI-Native Safety Intelligence Platform Co-Founded With The AES Corporation And AI Fund

Haven applies artificial intelligence to modernize incident investigations, root cause analysis, and prevention across high-risk industries

Haven Safety AI today announced its official launch. The company was co-founded in partnership with The AES Corporation, a global energy company operating in 12 countries and AI Fund, the venture studio founded by Andrew Ng. Haven is an AI-native platform designed to help organizations investigate incidents faster, uncover systemic risk, and prevent serious injuries before they occur.

Workplace safety remains one of the most operationally critical yet technologically underserved functions inside large enterprises. While traditional safety metrics and compliance programs helped reduce injuries, the industry has hit a serious injury and fatality (SIF) plateau, where serious incidents and fatalities persist despite years of investment in reporting and compliance. In addition to their irreparable impact on workers and their families, workplace injuries cost the U.S. economy more than $160 billion each year.

The company's platform acts as an AI copilot for safety teams. It guides structured witness statements, automatically gathers evidence, synthesizes timelines, identifies root causes, and recommends corrective actions grounded in both regulatory standards and an organization's historical data. Haven delivers measurable impact from day one by driving rapid gains in investigation efficiency, quality, and consistency; reducing repeat incidents; and enabling long-term improvements in risk exposure and serious incident reduction. Over time, the system develops institutional memory, enabling predictive insights and earlier intervention.

Haven is led by Joseph Hanna, a seasoned entrepreneur and product strategist with deep experience in leveraging AI to build operating systems for complex enterprise environments.

"Every serious incident leaves behind a trail of signals, but today those signals are scattered across interviews, documents, and disconnected systems," said Hanna, Co-Founder and CEO of Haven. "We built Haven to connect those dots automatically. Our goal is simple: help safety leaders understand risk sooner, act faster, and prevent the next incident rather than just documenting the last one."

"Safety investigations are critical and their outcomes affect human lives, regulatory compliance, and organizational trust. Yet they generate vast amounts of complex, unstructured data that most organizations still analyze manually," said Andrew Ng, Managing General Partner at AI Fund. "By combining modern AI with deep domain context, Joseph Hanna and the Haven team are on a trajectory to dramatically improve both the speed and quality of decision-making in high-consequence safety investigations."

The platform is organized into integrated modules that support the full lifecycle of safety management:

• havenSIGHT automatically collects witness statements, processes images, and frontline observations
• havenEDGE analyzes incidents, surfaces causal patterns, and recommends corrective actions
• havenIMPACT tracks outcomes over time and identifies leading indicators of future risk

Together, these capabilities transform safety from reactive reporting into proactive prevention.

"We are committed to using every tool we can to see our team members home safely at the end of every workday," said Chris Shelton, AES Chief Product Officer and President, AES Next. "We are excited to launch Haven through our partnership with AI Fund, turbocharging safety programs across electricity and other industries. Haven allows our crews and safety teams to quickly digest more insights than traditional tools on the market, going from incident, to cause, to corrective actions at unprecedented speed and scale."

Haven is initially focused on industries with complex operations and elevated risk profiles, including energy, utilities, construction, manufacturing, and logistics. After a year and a half of building, testing, and field validation, the platform is now live.