Google DeepMind
Gemini
Huge context window, strong analysis.
Impact Indicators
Job Displacement
Independent labor research finds no broad measurable disruption to employment from generative AI so far; no Gemini-specific outcome data exists.
Privacy & Surveillance
Google discloses that some Gemini conversations are reviewed by humans and kept up to three years when Gemini Apps Activity is on.
Water Usage
Google reports billions of gallons of water withdrawn yearly for data-center cooling; no Gemini-specific water figure is disclosed.
Energy & Carbon
IEA finds data-centre electricity demand surging worldwide, met partly by gas and coal; no independently verified Gemini-specific energy or emissions figure exists.
U.S. reporting requirements do not yet provide standardized, model-level disclosure for many AI human and environmental impacts. When U.S. evidence is incomplete, DoctorKnow uses verifiable EU regulatory disclosures, public-sector data, independent research, and reputable journalism to provide the strongest evidence currently available.
Evidence scope is identified as model-specific, provider-level, facility-level, regional/grid, or industry-level.
Scores use a 0–100 scale. Higher scores indicate stronger performance or lower concern; 100 is the best possible rating.
Reasoning
95
Coding
90
Writing
90
Analysis
96
Speed
70
Scores use a 0–100 scale, where 100 represents the strongest outcome and 0 the weakest.
What Gemini does
Gemini is built by Google DeepMind and is positioned for analysis, research and multimodal. Our readings score it strongest on data analysis and multi-step reasoning.
Its weakest measured axis is response speed (70/100), so plan around that when the work depends on it. Access is paid, and the weights are closed, so you can only reach it through the vendor's own service.
Ethical stance
Transparency reads 66/100. Some model cards and evaluations are public, but training data and methods are only partly described.
Labor and job-displacement posture reads 56/100. There is partial disclosure on data work and workforce impact, with clear gaps.
Regulatory posture reads 70/100. Gemini's vendor engages constructively with binding AI rules.
Privacy and data handling
Privacy and data handling reads 54/100. Training opt-out and retention controls exist, but defaults or consumer tiers are less protective.
Because the weights are closed, every prompt is processed on vendor infrastructure — treat regulated or confidential data accordingly.
Energy, carbon and water
Carbon footprint reads 74/100. Serving runs largely on low-carbon power, with credible reporting behind it.
Water and local impact reads 54/100. Some sites use evaporative cooling in regions with moderate water stress.
Speed and efficiency also matter here: a 70/100 speed reading means fewer compute-seconds per answer for routine work.
Diagnostic notes
Water reporting aggregated at fleet level, not per site.
Readings are DoctorKnow.ai's own 0–100 assessments, compiled from vendor documentation and public reporting, and revised as new disclosures land. They are a comparison aid, not a certification.