OpenAI commits $1 billion in Daybreak access to protect essential services

OpenAI has introduced Daybreak for Frontline Defenders, a global initiative meant to put its frontier cyber-defense capabilities into the hands of organizations that protect essential services but lack the budgets of large enterprises. The program combines a $1 billion global commitment to subsidized access to OpenAI's Daybreak cyber models and products, training, technical support and partnerships in the United States and internationally; a US-focused strand called Daybreak for America, which folds in a new pilot with the Multi-State Information Sharing and Analysis Center (MS-ISAC); and more than 35 enterprise products and partner-operated services delivered through the Daybreak Defense Network.

The $1 billion is targeted to be consumed within six months, starting with the United States. OpenAI says it will prioritize operators of essential services: water and wastewater systems, electric grid operators, state and local governments, community and regional banks, nonprofits and open-source maintainers, groups it describes as defending complex, often aging systems without the resources of the largest companies. Daybreak access is meant to help these teams review legacy code, analyze suspicious activity, identify and validate vulnerabilities, prioritize the most serious risks, and develop and test fixes. OpenAI intends to extend the model to partner countries in the coming weeks, though it does not name which ones.

The commitment builds on assistance OpenAI says it already provided after recent attacks on US water systems, when it offered affected states and utilities up to $1 million in no-cost API credits, Daybreak access and technical assistance; those teams used the support to review code and configurations, validate findings, develop patches and confirm fixes while the water systems stayed operational. Daybreak itself launched earlier this year in two tiers: Daybreak Blue, for common defensive work on OpenAI's mainline models, and Daybreak Red, which gives approved organizations access to specialized cyber models for more sensitive, technically demanding work. OpenAI says thousands of defenders across 2,000 approved organizations and workspaces, including cybersecurity companies, defense organizations and law enforcement agencies, already use Daybreak.

The new MS-ISAC pilot focuses on the public sector and water systems, pairing Daybreak access with guided training and hands-on assistance for an initial group of state, local, tribal and territorial defenders, with the aim of building a repeatable process that can later expand. MS-ISAC supplies cyber-threat intelligence, incident-response support and real-time information sharing to thousands of public-sector organizations, among them utilities, public hospitals, K-12 schools and law-enforcement agencies. Separately, OpenAI has been convening frontline defenders directly: this week's second gathering of utility companies brought together participants representing 40 states and the District of Columbia, whose utilities collectively serve more than half of the US population. OpenAI also says it published its approach to a Defense Factory, a continuous, agent-first operation built on existing security and engineering tools that finds and validates vulnerabilities and prepares tested fixes for review, and that it is sharing the architecture so other defenders can adapt it.

OpenAI frames the effort around what it calls a narrowing defender's window, the period in which AI can help defenders close security gaps before attackers exploit increasingly capable models at scale; it warns that in the coming months AI-enabled cyber attacks will become far more widespread and sophisticated as models around the world grow more capable. Last week it joined more than 150 organizations across cybersecurity, technology, critical infrastructure, finance and AI in a call for collective action, arguing that no single company can secure the systems everyone depends on. Eligible state and local governments, critical infrastructure operators, nonprofits, open-source maintainers and other supporting organizations can apply for access, training and technical assistance through the Daybreak website.

Key facts

OpenAI is committing $1 billion globally to subsidized Daybreak cyber-model access, training, technical support and partnerships, targeted for use within six months, starting with the United States.
The program prioritizes water and wastewater systems, electric grid operators, state and local governments, community and regional banks, nonprofits and open-source maintainers.
A new pilot with the Multi-State Information Sharing and Analysis Center (MS-ISAC) pairs Daybreak access with guided training for state, local, tribal and territorial defenders, focused on the public sector and water systems.
More than 35 enterprise products and partner-operated services now bring Daybreak's cyber models into existing workflows through the Daybreak Defense Network.
Daybreak already has thousands of defenders across 2,000 approved organizations, split between Daybreak Blue for common defensive work and Daybreak Red for specialized, more sensitive cyber work.

Why it matters

Frontline defenders such as water utilities, electric grid operators, local governments and community banks are responsible for systems many people depend on every day, but they often lack the staff, budgets and specialized security expertise of large enterprises. OpenAI argues that AI is simultaneously making attacks against those systems more capable and, used correctly, giving defenders a way to close gaps faster than attackers can exploit them; it calls this a narrowing defender's window. That framing is also why OpenAI positions the effort as a shared responsibility rather than a product pitch alone: it says no single company can secure the systems everyone depends on, and it joined more than 150 other organizations last week in calling for coordinated action across the cybersecurity, technology, critical infrastructure and finance sectors.

Who it affects

OpenAI names specific categories of intended beneficiaries: water and wastewater systems, electric grid operators, state and local governments, community and regional banks, nonprofits and open-source maintainers. The MS-ISAC pilot extends that same public-sector and water focus to state, local, tribal and territorial defenders; MS-ISAC's broader membership includes utilities, public hospitals, K-12 schools and law-enforcement agencies. OpenAI also points to direct engagement: this week's second gathering of utility companies brought together participants representing 40 states and the District of Columbia, whose utilities collectively serve more than half of the US population. Internationally, OpenAI intends to extend the model to partner countries in the coming weeks, though none are named yet.

How to use it

Eligible state and local governments, critical infrastructure operators, nonprofits, open-source maintainers and other supporting organizations can apply through the Daybreak website for subsidized access, training and technical assistance. Access comes in two tiers OpenAI introduced earlier this year: Daybreak Blue, built on its mainline models for common defensive work, and Daybreak Red, which gives approved organizations specialized cyber models for more sensitive, technically demanding work. OpenAI describes practical uses that include reviewing legacy code, analyzing suspicious activity, identifying and validating vulnerabilities, prioritizing the most serious risks, and developing and testing fixes. The source does not state a discount percentage or a specific dollar figure per organization for what counts as subsidized access, so no per-user pricing can be given here.

How solid is it

This account comes entirely from OpenAI's own announcement, so the figures are self-reported and not independently verified. Some claims come with a track record attached: OpenAI says Daybreak already has thousands of defenders across 2,000 approved organizations and workspaces, including cybersecurity companies, defense organizations and law-enforcement agencies, and that its prior support to states and utilities after recent attacks on US water systems, up to $1 million in no-cost API credits, Daybreak access and technical assistance, was already used to review code and configurations, validate findings, develop patches and confirm fixes while the affected systems stayed operational. The utility gathering's count of 40 states plus the District of Columbia, serving more than half the US population, and the more than 150 organizations behind last week's call for collective action are concrete, checkable figures rather than vague claims. What is missing is any external confirmation of the $1 billion figure itself or of how it will actually be spent.

Risks and caveats

The source does not break down how the $1 billion splits between subsidized access, training, technical support and partnerships, or how it divides between the United States and other countries. No named executive or spokesperson is quoted; every statement is attributed to OpenAI collectively. The announcement does not identify the specific water-system attacks it references, nor does it list the more than 35 partner products and services or the companies behind them. No discount percentage or per-organization dollar value is given for what OpenAI calls subsidized access, and no partner countries are named for the planned international expansion. Readers should treat the program as OpenAI's own framing of its cyber-defense work until independent reporting or user accounts test the claims.

“We have a defender's window: a narrowing opportunity to use AI to close security gaps before attackers seize them.”

— OpenAI