OpenAI has raised the rewards in its Security Bug Bounty program. In a post dated November 2, 2026, OpenAI said: "A critical vulnerability in a core product such as ChatGPT, Codex, or the API Platform will now earn up to $30,000." It added: "Exceptional critical findings on those targets remain eligible for up to $100,000, and now up to $172,000 with bonuses."

OpenAI said the aim is "to further encourage independent researchers to pursue vulnerabilities in OpenAI software that our testing may have missed." The same day, it said in a second post that it is partnering with Cathedral for its next internal security sprint.

#How much are rewards going up?

OpenAI's bounty post puts it this way: "We’re more than quadrupling the midpoint of our reward ranges for medium, high, and critical findings on average." Its sprint post describes the change as "more than quadrupling our rewards for critical findings."

The $100,000 ceiling for exceptional findings is not new. OpenAI says those findings "remain eligible" for that amount. The higher total comes from bonuses stacked on top of the base bounty.

#How do the bonuses work?

OpenAI lists three kinds:

  • Quality. "High-quality reports will be eligible to earn a bonus multiplier on the base bounty."
  • Loyalty. "This 5-tier system provides scaling bonuses to recognize, support, and celebrate researchers who consistently partner with us in making AI technology safer for everyone."
  • Priority campaigns. OpenAI says "priority campaigns will feature dedicated bonus pools for newly launched features".

OpenAI gives worked examples. A critical zero-click account takeover in ChatGPT, which it lists at $6,500 historically, would earn $51,577 under the new system with all three bonuses applied to a $30,000 base. An exceptional finding at the $100,000 base with the same bonuses totals $171,925. OpenAI notes: "All scenarios and historical awards are illustrative."

The Decoder wrote that the change "puts OpenAI in the same ballpark as Google's general vulnerability rewards program, though still below the top-tier payouts Apple or Microsoft offer for OS-level exploits." That comparison is The Decoder's, not OpenAI's.

#Do earlier reports get the new rates?

OpenAI sets two rules:

  1. Submitted, unawarded. "If you’ve already submitted a report and we haven’t yet awarded a bounty, we’ll apply the new reward ranges."
  2. Recently awarded. "For bounties we awarded in the 30 days before this announcement, we’ll pay you the difference if your finding would have earned more under the new ranges (Base Award + Loyalty Awards)."

The top-up applies only to bounties OpenAI awarded in the 30 days before its November 2, 2026 announcement.

OpenAI also said its security team will be at PROMPT2PWN in Oakland on November 20 and 21, 2026, "where researchers can test targets in person."

#What is the Cathedral security sprint?

"We’re planning our next internal security sprint at OpenAI, bringing new researchers and security partners together with our own teams to use our models to test and strengthen our systems," OpenAI wrote.

On Cathedral, OpenAI said: "We’re partnering with Cathedral, whose researchers will use OpenAI’s models alongside our teams to challenge our defenses, bringing differentiated expertise and approaches to finding weaknesses and helping us protect our platforms." OpenAI has not given a start date for the sprint.

OpenAI's post quotes Gavin Kliger, CEO and co-founder of Cathedral: "OpenAI is setting an important example by partnering with outside security firms and using its own models to continuously test and strengthen its defenses." He added: "We’re excited to collaborate with OpenAI’s teams, combining their models and security expertise with our research to find weaknesses, test fixes and strengthen this approach to continuous defense."

#What happened in the previous sprint?

OpenAI says that in its previous sprint, "We dedicated 25% of our product engineers to finding and fixing weaknesses in our systems." It adds: "Using Codex, engineers investigated findings, reproduced vulnerabilities in isolated environments and prepared patches for review."

OpenAI frames the work as necessary "because traditional cyber defenses alone are no longer sufficient." The bounty changes complement what OpenAI calls its Defense Factory work, "where our engineers use our models to find weaknesses, reproduce them, and test fixes".

OpenAI also says it is making its most capable models available to Daybreak customers, and: "We’ve committed $1 billion in subsidized access over six months for frontline defenders protecting essential services and digital infrastructure."

For more on OpenAI's plans, see our reports on OpenAI's Grid Protectors Program and OpenAI's Pro usage limits.

Our analysis: The headline $172,000 is a ceiling that requires every bonus at once, as OpenAI's own worked example shows. The base cap for exceptional findings stays at $100,000. The bigger shift is for ordinary critical bugs: OpenAI's example shows a core-product account takeover rising from $6,500 to $51,577 with bonuses. Pairing higher bounties with an outside sprint partner signals that OpenAI wants more external eyes on its products alongside its own model-driven testing.

#What is unclear?

  • Full ranges. OpenAI's posts give caps and examples, not the complete new reward table.
  • Sprint timing. OpenAI gives no start date for the Cathedral sprint.