Google's Gemini 4 Carbon aims to own the AI cyber defense lifecycle

Gemini 4 Carbon debuts

Google announced Gemini 4 Carbon on October 28, 2026 as its latest frontier model, emphasizing cybersecurity defense and long-horizon engineering tasks. Koray Kavukcuoglu, Google DeepMind senior vice president and chief AI architect, said Carbon ranks first on Artificial Analysis Intelligence Index v5 at score 59, extending gains from Gemini 4 Argon released weeks earlier.

Google is taking a phased rollout: initial access for a subset of cyber defenders, then expansion through the Fairwind Program that already includes more than 650 organizations such as governments and critical infrastructure operators.

Benchmarks across find, exploit, patch

Google reported 88.2% on SEC-bench Pro for vulnerability discovery, 54.8% on ExploitGym for exploitation validation, and 71.0% on Collinear's CWE-bench for patching—leading comparisons that included OpenAI GPT-6 Astra and Anthropic Claude Opus 5.5 and Claude Fable 5.1 in Google's charts. Results were measured without guardrails, the company noted.

Internal benchmarks claimed more than 2x secure-code generation success versus Argon; Chrome engineers saw Carbon beat a rival frontier model by more than 10x on high-quality patches. Wiz observed leadership on web penetration testing benchmarks tied to its Scan for Good initiative.

Safety investments

Carbon builds on Argon's multi-layer safety stack—training, misuse defenses, prompt-injection resilience, and chain-of-thought monitoring. Google highlighted reduced shortcut exploitation and fewer unsupported claims versus Argon, plus zero successful indirect prompt injection attacks in Gray Swan testing where Argon saw 0.7% success.

Reuters quoted introductory pricing at $2 per million input tokens and $10 per million output tokens, matching Argon's launch offer before rising to $4/$20.

Fairwind and government testing

Google said it engaged the U.S. government's voluntary pre-release model testing process and will iterate guardrails using defender feedback before consumer and developer GA.

SecurityWeek emphasized Carbon is not broadly available yet, a constraint defenders accept if early access accelerates patch pipelines on real codebases.

Implications for enterprise security

CISOs should treat Carbon as a force multiplier for AppSec triage—not autonomous patch deployment without human review. The model's exploit benchmarks mean red teams must assume attackers will eventually access similar capabilities via stolen keys or jailbreaks.

Google's cadence—Argon then Carbon within weeks—signals annualized frontier releases becoming quarterly, compressing vendor evaluation cycles.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.

Pricing parity with Argon may accelerate enterprise trials once GA arrives.

Restricted rollouts aim to keep exploit-capable models away from opportunistic abuse while defenders gain experience.

Benchmark leadership without guardrails does not automatically translate to safe default product settings.