Transparency problem of big AI Models Revealed by Stanford Study

Lack of Transparency in Key AI Models Found by Stanford Study

[{"selector":"#anim-f1cfe239-5336-4803-afae-e01b3aec8e88","keyframes":{"transform":["translate3d(-135.41667%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-e17fa0b8-7f5c-485b-a215-6f1ba0b18893","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-3df84ab5-3102-4e6b-bf1c-8ba70fdc131a","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Stanford study evaluates top AI models' transparency across 100 indicators

[{"selector":"#anim-f34e8ebe-c6aa-4e56-a859-ca3b93becc8a","keyframes":{"transform":["translate3d(-125.64102%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-4c6482a4-5c07-4440-9395-f3397974497d","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-1a93c773-23e5-49a7-aa66-6c226e985eda","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Results showed that none of the models had a transparency index score above 54%, and a few had a score as low as 12%.

[{"selector":"#anim-065adc32-eaf9-4908-b540-795801c5bc31","keyframes":{"transform":["translate3d(-120.76678%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-3c730fba-cf40-4155-bef5-e476385e9c2d","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-180c79cd-2456-4f41-a04b-61246d085529","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Researchers at Stanford Warn of Risks Due to Lack of AI Transparency

[{"selector":"#anim-76a6ad84-4ec5-40f1-8442-9f0cce9945d9","keyframes":{"transform":["translate3d(-126.48222%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-6a5beda4-e253-41f0-acfc-f5716d930dbf","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-71aaeb84-c254-4f2f-b772-51955199d02a","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

They exhorted the firms that built the models to give more details and embrace best practices for transparency.

[{"selector":"#anim-2ec5492f-f040-4997-9789-783977eb8a2d","keyframes":{"transform":["translate3d(-122.10526%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-7f7951bf-b7a5-4fc4-ad4e-ef91cb189a41","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-3e35a788-9c49-42c5-8000-93319f827edf","keyframes":{"transform":["scale(0.15)","scale(1)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"forwards"}]

Transparency problem of big AI Models Revealed by Stanford Study

Lack of Transparency in Key AI Models Found by Stanford Study

Stanford study evaluates top AI models' transparency across 100 indicators

Results showed that none of the models had a transparency index score above 54%, and a few had a score as low as 12%.

Researchers at Stanford Warn of Risks Due to Lack of AI Transparency

They exhorted the firms that built the models to give more details and embrace best practices for transparency.

Read More Stories