AI Safety
What AI providers state their models reached in safety testing, in their own words, and what outside testers found.
Also: compare models side by side · statements that a capability level was crossed · providers · practices providers document · safety news.
What providers state
Each row is one statement in a provider's own safety document about a capability level of one of its models: the level as the provider names it, whether the provider says the model reached it, and the sentence that says so.
139 of 139
| Model | Level, as the provider names it | The provider states | The provider's sentence |
|---|---|---|---|
| Muse Spark Meta |
Moderate or lower Autonomy and AI research |
Not reached | “We assess loss of control risk as “moderate or lower”.” source |
| Muse Spark Meta |
High Chemical, biological, radiological, nuclear |
Not determined | “Those evaluations identified elevated risks prior to mitigations, with Chemical and Biological capabilities assessed as likely reaching the “high risk” category under the Advanced AI Scaling Framework before safeguards were applied.” source |
| Muse Spark Meta |
Moderate or lower Chemical, biological, radiological, nuclear |
Not reached | “However, after applying mitigations, the residual risk of the model is that of “moderate or lower risk.”” source |
| Muse Spark Meta |
Below the level required to meaningfully enable the cyber threat scenarios defin Cyber |
Not reached | “Muse Spark’s offensive cyber capabilities remain below the level required to meaningfully enable the cyber threat scenarios defined in the Advanced AI Scaling Framework.” source |
| Muse Spark Meta |
Moderate or lower Cyber |
Not reached | “We assess cybersecurity risk as “moderate or lower” with respect to the threat scenarios outlined in Meta’s Advanced AI Scaling Framework.” source |
| Gemini 4 Argon Google DeepMind |
TCL (Tracked Capability Level) Autonomy and AI research |
Not determined | “In light of this, we assess that we cannot rule out the TCL.” source |
| Gemini 4 Argon Google DeepMind |
TCL (Tracked Capability Level) Chemical, biological, radiological, nuclear |
Not reached | “We can rule out the TCL for the chemical, radiological, and nuclear domains with reasonable confidence.” source |
| Gemini 4 Argon Google DeepMind |
CCL (Critical Capability Level) Deception and sabotage |
Not reached | “we assess its overall capabilities to be below the CCL threshold.” source |
| Gemini 4 Argon Google DeepMind |
CCL (Critical Capability Level) Misuse and jailbreaks |
Not reached | “Gemini 4 Argon demonstrates some ability to influence user beliefs and behaviors during one-on-one direct conversations in human behavioral studies, and may provide utility in the planning stage of a malicious influence operation but we assess its overall capabilities to be below the CCL threshold.” source |
| Gemini 4 Argon Google DeepMind |
CCL (Critical Capability Level) Other |
Not reached | “Gemini 4 Argon's performance remains below our CCL threshold.” source |
| Gemini 3.7 Flash Google DeepMind |
CCL (Critical Capability Level) Autonomy and AI research |
Not reached | “The model does not reach the CCL alert threshold.” source |
| Gemini 3.7 Flash Google DeepMind |
CCL (Critical Capability Level) Chemical, biological, radiological, nuclear |
Not reached | “We can rule out the CCL for the CBRN domain with reasonable confidence based on the results from our testing.” source |
| Gemini 3.7 Flash Google DeepMind |
TCL (Tracked Capability Level) Chemical, biological, radiological, nuclear |
Not reached | “We can rule out the TCL for the CBRN domain with reasonable confidence based on the results from our testing.” source |
| Gemini 3.7 Flash Google DeepMind |
CCL (Critical Capability Level) Cyber |
Not reached | “Gemini 3.7 Flash reaches the alert threshold for this CCL, but not the CCL.” source |
| Gemini 3.7 Flash Google DeepMind |
TCL (Tracked Capability Level) Deception and sabotage |
Not reached | “The model does not reach the TCL.” source |
| Gemini 3.5 Transcribe Live Google DeepMind |
CCL (Critical Capability Level) Other |
Not determined | “Our assessments have shown that none of Gemini 3.5 Translate, Gemini 3.5 Transcribe, or Gemini 3.5 Transcribe Live have meaningful new T/CCL-relevant capabilities or material increases in performance with respect to Frontier Safety compared to Gemini 3.1 Pro and Gemini 3.7 Flash, therefore we are confident that none of them are likely to reach any CCLs.” source |
| Gemini 3.5 Transcribe Google DeepMind |
CCL (Critical Capability Level) Other |
Not determined | “Our assessments have shown that none of Gemini 3.5 Translate, Gemini 3.5 Transcribe, or Gemini 3.5 Transcribe Live have meaningful new T/CCL-relevant capabilities or material increases in performance with respect to Frontier Safety compared to Gemini 3.1 Pro and Gemini 3.7 Flash, therefore we are confident that none of them are likely to reach any CCLs.” source |
| Gemini 3.5 Flash-Lite Google DeepMind |
CCL (Critical Capability Level) Other |
Not determined | “Our assessments have shown that, while Gemini 3.5 Flash-Lite excels at agents and coding, it does not have meaningful new capabilities or material increases in performance with respect to the domains outlined in our Frontier Safety Framework compared to Gemini 3.1 Pro; therefore, based on Gemini 3.1 Pro results, we are confident that Gemini 3.5 Flash-Lite is also unlikely to reach any CCLs.” source |
| Gemini 3.5 Audio Google DeepMind |
Tracked or critical capability level (t/ccls) Other |
Not reached | “they did not reach any Tracked or Critical Capability Levels (T/CCLs) outlined in our Frontier Safety Framework .” source |
| Gemini 3.1 Pro Google DeepMind |
Uplift level 1 Chemical, biological, radiological, nuclear |
Not reached | “We found that the model remains below alert thresholds for the CBRN, harmful manipulation, machine learning R&D, and misalignment CCLs.” source |
| Gemini 3.1 Pro Google DeepMind |
Uplift level 1 Cyber |
Not reached | “As previous models passed the alert threshold for cyber, we performed more additional testing in this domain on Gemini 3.1 Pro with and without Deep Think mode, and found that the model remains below the cyber CCL.” source |
| Gemini 3.1 Flash-Lite Google DeepMind |
CCL (Critical Capability Level) Other |
Not determined | “Our assessments have shown that Gemini 3.1 Flash-Lite is less capable than Gemini 3.1 Pro, therefore based on Gemini 3.1 Pro, we are confident that Gemini 3.1 Flash-Lite is also unlikely to reach any CCLs.” source |
| Gemini 3.1 Flash Audio Google DeepMind |
CCL (Critical Capability Level) Other |
Not determined | “Our assessments have shown that Gemini 3.1 Flash Audio is less capable than Gemini 3.1 Pro.” source |
| Gemini 2.5 Pro Google DeepMind |
Machine Learning R&D Uplift Level 1 Autonomy and AI research |
Not determined | “given Gemini 2.5 Pro Preview's poor median performance on RE-Bench relative to experts, the model likely lacks the necessary capabilities to automate or significantly uplift any significant fraction of the research process.” source |
| Gemini 2.5 Pro Google DeepMind |
Instrumental Reasoning Level 1 Deception and sabotage |
Not reached | “Instru- mental Reasoning Level 1 ✘ CCL not reached” source |
| Gemini 2.5 Pro Google DeepMind |
Instrumental Reasoning Level 2 Deception and sabotage |
Not reached | “Instru- mental Reasoning Level 2 ✘ CCL not reached” source |
| Gemini 2.5 Deep Think Google DeepMind |
Uplift level 1 Chemical, biological, radiological, nuclear |
Not determined | “Our assessment is that the model has enough technical knowledge in certain CBRN scenarios and stages to be considered at early warning.” source |
| Gemini 2.5 Deep Think Google DeepMind |
Cyber Uplift Level 1 Cyber |
Not determined | “CCL reached? No, but the early warning alert threshold was originally reached by Gemini 2.5 Pro and continues to be met by Gemini 2.5 Deep Think.” source |
| Gemini 2.5 Deep Think Google DeepMind |
Uplift level 1 Cyber |
Not determined | “For Cyber Uplift Level 1 , the early warning alert threshold was originally reached by Gemini 2.5 Pro and continues to be met by Gemini 2.5 Deep Think.” source |
| GPT-6.1 Terra OpenAI |
High Autonomy and AI research |
Not reached | “In AI Self-Improvement, GPT-6.1 Terra does not reach our High threshold.” source |
| GPT-6.1 Terra OpenAI |
High Chemical, biological, radiological, nuclear |
Reached | “Under our Preparedness Framework, we are treating GPT-6.1 Terra as Critical in cybersecurity and High for Biological and Chemical capability.” source |
| GPT-6.1 Terra OpenAI |
Critical Cyber |
Reached | “Under our Preparedness Framework, we are treating GPT-6.1 Terra as Critical in cybersecurity and High for Biological and Chemical capability.” source |
| GPT-5.5 Instant OpenAI |
Below high capability Autonomy and AI research |
Not reached | “GPT-5.5 Instant is considered below High Capability.” source |
| GPT-5.5 Instant OpenAI |
High Autonomy and AI research |
Not reached | “For AI Self-Improvement, evaluations of final checkpoints indicate that, like its predecessor models, GPT-5.5 Instant does not have a plausible chance of reaching a High threshold.” source |
| GPT-5.5 Instant OpenAI |
High Chemical, biological, radiological, nuclear |
Reached | “GPT-5.5 Instant is our first Instant model to be treated as High Capability in the Biological and Chemical domain, as well as in the Cybersecurity domain.” source |
| GPT-5.5 Instant OpenAI |
High Cyber |
Reached | “this is the first Instant model that we are treating as High capability in our Cybersecurity and Biological & Chemical Preparedness categories, and implementing appropriate safeguards.” source |
| GPT-5 OpenAI |
High Chemical, biological, radiological, nuclear |
Reached | “Similarly to ChatGPT agent, we have decided to treat gpt-5-thinking as High capability in the Biological and Chemical domain under our Preparedness Framework, activating the associated safeguards.” source |
| GPT-5 OpenAI |
High cyber risk Cyber |
Not reached | “The gpt-5 model series does not meet the threshold for high cyber risk.” source |
| Claude Sonnet 4.5 Anthropic |
AI R&D-4 Autonomy and AI research |
Not reached | “All models tested remain below our pre-committed rule-out threshold of 0.6.” source |
| Claude Sonnet 4.5 Anthropic |
ASL-3 Autonomy and AI research |
Not determined | “Exhibited improved autonomy capabilities in software engineering and AI research tasks, though still below the ASL-4 rule-out threshold.” source |
What outside testers report
Each row is one result from a government evaluator, independent lab, academic or security firm that tested a model, quoted from where it is reported.
1940 of 1940
| Model | Tester | Test | What was found |
|---|---|---|---|
| o3 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
physical validation proof-of-concept Chemical, biological, radiological, nuclear |
“They then successfully followed o3’s protocol, with the minor variations described in Appendix C, to assemble those segments into a valid eGFP gene, insert that gene into a plasmid, and express it to confirm that the protein was functional.” source |
| o3 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
egfp-targeted task Chemical, biological, radiological, nuclear |
“OpenAI’s o3 and Claude Opus 4 both performed well when attempting to create eGFP, but both denied approximately half of the task requests.” source |
| o3 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
influenza ha oligonucleotide synthesis task Chemical, biological, radiological, nuclear |
“All four of those models refused every request to create oligonucleotides for the production of influenza HA that we issued during this project.” source |
| o3 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
polymerase chain assembly (PCA) task Chemical, biological, radiological, nuclear |
“sk attempt by an o3-driven agent demonstrated the ability to produce biologically coherent sequences and protocols for a PCA task.” source |
| o3 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
physical validation of o3-designed DNA sequences and o3-generated laboratory protocol Chemical, biological, radiological, nuclear |
“The validation experiment showed that o3-designed DNA sequences could be assembled into a functional eGFP protein and that the o3-generated protocol was accurate.” source |
| o3 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
egfp task Chemical, biological, radiological, nuclear |
“o3 refused 6/10 requests for eGFP, so it generated significantly fewer performance data than many of our other models.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
benchtop DNA synthesizer agent evaluation Chemical, biological, radiological, nuclear |
“The most recent models we tested, GPT-5, Opus 4.5, and Gemini 3 Pro, reliably design eGFP DNA segments that satisfy all our scoring criteria including expert review.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
EGFP task Chemical, biological, radiological, nuclear |
“Figure D.9. eGFP Task Performance Results, Gemini 3 Pro” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
Influenza A HA segment 4 task Chemical, biological, radiological, nuclear |
“Figure D.10. Influenza A HA Segment 4 Task Performance Results, Gemini 3 Pro” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
synthesizer-based DNA acquisition task Chemical, biological, radiological, nuclear |
“The DNA acquisition task described in this report tests a simple and widely accessible agent framework for its ability to submit an appropriate set of DNA sequences to a simulated benchtop synthesizer interface.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
benchtop DNA synthesis task Chemical, biological, radiological, nuclear |
“For our assessment, we create an open-ended test environment and present agents with our synthesis task.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
DNA acquisition task Chemical, biological, radiological, nuclear |
“Though this means that mechanics of the sequence retrieval step differed between those two models and the remainder of our set, it was demonstrated that neither Opus 4.5 nor Gemini 3 Pro needed database search tools to succeed at our task (see Chapter 5 for test results).” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
agent task attempt Chemical, biological, radiological, nuclear |
“Each successful agent task attempt generated a set of DNA segments that we graded against a number of feasibility criteria.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
segment scorer Chemical, biological, radiological, nuclear |
“We implemented the segment scorer as a set of rule-based checks for five criteria, which are listed here but described more comprehensively in Appendix B.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
llm transcript autograding Chemical, biological, radiological, nuclear |
“Our LLM transcript autograding was used to create additional grading dependencies, as described in the next section.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
manual segment review Chemical, biological, radiological, nuclear |
“For further validation of each submission that received a score of 5/5 on the above criteria, we referred it to one of our team’s biology experts for additional review.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
laboratory protocol autograding Chemical, biological, radiological, nuclear |
“The same autograding system used to evaluate our conceptual checks on agent trajectories was used to evaluate the agent-generated laboratory protocols that accompanied submissions of segments that successfully encoded the target DNA sequence.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
egfp-targeted task Chemical, biological, radiological, nuclear |
“To generate the data reported in the next chapter, each model was prompted to attempt our task ten times for each of two protein targets: eGFP and influenza HA.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
egfp DNA sequence design task with simulated synthesizer and protocol autograding Chemical, biological, radiological, nuclear |
“Several months later, Claude Opus 4.5 and Gemini 3 Pro achieved the same milestone.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
influenza HA oligonucleotide design task with simulated synthesizer Chemical, biological, radiological, nuclear |
“Gemini 3 Pro was the only model that was both capable of regularly producing acceptable oligonucleotides and that did so for the influenza HA target.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
attachment of appropriate restriction enzyme sites to designed segments Chemical, biological, radiological, nuclear |
“This was not a problem for Gemini 3 Pro, however, which attached BsmBI to both ends of the insert sequence in all but one case.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
physical validation experiment Chemical, biological, radiological, nuclear |
“That validation resulted in the assembly and cloning of a complete gene and a functional eGFP protein.” source |
| Gemini 3 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
pca task Chemical, biological, radiological, nuclear |
“sk attempt by an o3-driven agent demonstrated the ability to produce biologically coherent sequences and protocols for a PCA task.” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
agent evaluation designed to probe capabilities in applied molecular biology Chemical, biological, radiological, nuclear |
“In this report, we describe the design and execution of an agent evaluation designed to probe capabilities in applied molecular biology.” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
eGFP task Chemical, biological, radiological, nuclear |
“Figure D.11. eGFP Task Performance Results, Gemini 2.5 Pro” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
influenza A HA segment 4 task Chemical, biological, radiological, nuclear |
“Figure D.12. Influenza A HA Segment 4 Task Performance Results, Gemini 2.5 Pro” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
DNA acquisition task Chemical, biological, radiological, nuclear |
“For our assessment, we create an open-ended test environment and present agents with our synthesis task.” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
DNA acquisition task (synthesis of EGFP and influenza A HA segment) Chemical, biological, radiological, nuclear |
“We used it to test a set of eight models within simple agent scaffolds: GPT-4.1, o3, and GPT-5 from OpenAI; Claude Sonnet 4, Opus 4, and Opus 4.5 from Anthropic; and Gemini 2.5 Pro and Gemini 3 Pro from Google.” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
protocol autograder Chemical, biological, radiological, nuclear |
“We used our autograder to assess 58 protocols, resulting in 290 autograded question-protocol pairs.81” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
egfp-targeted task Chemical, biological, radiological, nuclear |
“During evaluation development, and prior to generating the data reported in the next chapter, we used our eGFP-targeted task to adjust our ReAct implementation for each model.84” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
egfp protocol autograder Chemical, biological, radiological, nuclear |
“Following a misstep in the eGFP task, one of Gemini 2.5 Pro’s laboratory protocols incorrectly proposed Gibson assembly, rather than PCA, and did not mention RE digest for either insert or plasmid.92” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
DNA segment design task for EGFP and influenza A HA segment assembly Chemical, biological, radiological, nuclear |
“We found that models from the most recent generation of LLMs, beginning with GPT-5, consistently perform our task in ways that met all of our scoring criteria.” source |
| Gemini 2.5 Pro Google DeepMind |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
physical validation experiment of an o3-driven agent's PCA task attempt Chemical, biological, radiological, nuclear |
“We also describe a physical validation experiment in which a successful task attempt by an o3-driven agent demonstrated the ability to produce biologically coherent sequences and protocols for a PCA task.” source |
| GPT-5 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
agent evaluation designed to probe capabilities in applied molecular biology Chemical, biological, radiological, nuclear |
“The most recent models we tested, GPT-5, Opus 4.5, and Gemini 3 Pro, reliably design eGFP DNA segments that satisfy all our scoring criteria including expert review.” source |
| GPT-5 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
requests to aid in laboratory tasks Chemical, biological, radiological, nuclear |
“Claude Opus 4, Opus 4.5, o3, and GPT-5 denied all requests related to assisting with acquisition of our HA sequence, and several requests for eGFP as well.” source |
| GPT-5 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
EGFP task Chemical, biological, radiological, nuclear |
“Figure D.1. eGFP Task Performance Results, GPT-5 .................................................................. 57” source |
| GPT-5 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
synthesizer-based DNA acquisition task Chemical, biological, radiological, nuclear |
“In Chapter 4, we describe an evaluation involving a synthesizer- based DNA acquisition task, with a description of our implementation, scoring, and testing.” source |
| GPT-5 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
DNA synthesis task Chemical, biological, radiological, nuclear |
“For our assessment, we create an open-ended test environment and present agents with our synthesis task.” source |
| GPT-5 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
LLM autograders on transcripts of agent task attempts Chemical, biological, radiological, nuclear |
“As a supplement to our checks of agent-generated DNA sequences, we also make use of LLM autograders to provide feedback on transcripts of agent task attempts.” source |
| GPT-5 OpenAI |
RAND (CAST / Project Canary) Independent lab · 2026-02-03 |
DNA acquisition task Chemical, biological, radiological, nuclear |
“We used it to test a set of eight models within simple agent scaffolds: GPT-4.1, o3, and GPT-5 from OpenAI; Claude Sonnet 4, Opus 4, and Opus 4.5 from Anthropic; and Gemini 2.5 Pro and Gemini 3 Pro from Google.” source |