Gemma 4 / Ethics & safety

Gemma 4 Ethics & Safety

Google filtered Gemma 4's training data and tested the finished models for harmful output before release. But an open model you download arrives without the safety layers a hosted service wraps around it. This page explains what Google did, what is left to you, and how EmbeddingGemma 2 differs.

A safe deployment has layers Personprompt Input check Gemma 4safety-tested Output check Harmful request stopped you add this Google's part you add this
The model is one layer. The checks around it are yours to add.
Before release

What Google did

From the Gemma 4 model card: how the training data was cleaned and how the models were tested.

🚫

Child-safety filtering

Filtering for child sexual abuse material was applied at multiple stages of data preparation.

🔏

Sensitive data filtering

Automated methods removed certain personal information and other sensitive data from the training sets.

🧹

Quality and safety filtering

Further filtering based on content quality and safety, in line with Google's policies.

What was tested

Automated and human evaluations, run with Google's safety teams, aimed at preventing child exploitation material, dangerous content, sexually explicit content, hate speech and harassment. Testing covered text and images, with safety filters switched off.

What Google reports

"Gemma 4 models significantly outperform Gemma 3 and 3n models in improving safety, while keeping unjustified refusals low." The card describes minimal policy violations across all sizes. It gives no detailed figures.

From Google's Gemma 4 model card. These are the maker's own evaluations; this site has not tested the models' safety behaviour.

Named risks

The risks the model card names

⚖️

Bias and fairness

Models trained on large amounts of real-world material can reflect the social and cultural biases in it.

📰

Misinformation

The model can state things that are wrong or out of date, and could be misused to produce misleading text. Its knowledge stops at January 2025.

🔐

Privacy

Training data was filtered for personal information. Filtering lowers the risk; it is not a guarantee.

🛡️

Harmful content and misuse

Google's guidance is for developers to add content safety safeguards suited to their own product.

Paraphrased from the "Usage and limitations" section of the model card. What the models get wrong in ordinary use is on the limitations page.

The core difference

What open weights change

The same model behaves the same way wherever it runs. What changes is who is responsible for everything around it.

With a hosted service

  • The provider adds filters on what goes in and what comes out.
  • The provider watches for abuse and can cut off bad actors.
  • Updates arrive automatically.
  • Your prompts leave your machine and are handled under the provider's terms.

With open weights you run yourself

  • No filters unless you add them. The model's own training is the only safeguard.
  • Monitoring is yours to build, or to go without.
  • You choose when to update, and must remember to.
  • Fine-tuning can change safety behaviour, so test again afterwards.

Where open weights are the better choice

Prompts never leave the device

Run locally, nothing is sent to anyone. For sensitive material that can be the safest option there is. See run locally.

Open to inspection

Researchers can examine and test the actual model, and results can be reproduced because the weights do not change underneath you.

Access for everyone

No account, approval or payment is needed, which matters for education, research and places with poor connectivity.

This section is this site's view. Hosted options are on the API page; running a service yourself is under self-hosting.

Rules

The licence, and Google's use policies

DocumentWhat it isLink
Apache 2.0 licenceGemma 4's licence. Allows commercial use, changes and redistribution. Earlier Gemma versions used Google's own Gemma Terms of Use instead.Gemma 4 licence
Prohibited Use PolicyGoogle's list of uses it does not permit for Gemma models, linked from the Gemma 4 documentation.Prohibited use
Intended Use StatementGoogle's description of what the models are meant for.Intended use

What the Prohibited Use Policy covers

Violating others' rights

Content that infringes or misappropriates a person's or organisation's rights.

Dangerous, illegal or malicious activity

Including child exploitation, crime, violent extremism, spam, fraud, phishing, malware and getting around safety filters.

Harm to people or groups

Hatred, harassment, violence, self-harm, doxxing, surveillance and discrimination.

Misleading content

Impersonation, false claims of expertise, defamation, and automated decisions in high-stakes areas.

Sexually explicit content

For pornography or gratification. Scientific, educational, documentary and artistic uses are excepted.

Not legal advice

How the policy and the Apache 2.0 licence fit together for your use is a question for a lawyer, not this page.

Summarised from Google's Gemma Prohibited Use Policy; read the original for the exact wording. Licence differences between versions are on the Gemma 3 page and in the model comparison.

Google's recommended process

Design, align, evaluate, protect

The four stages of Google's Responsible Generative AI Toolkit, with what each means in practice.

  1. Design

    Decide what your application should and should not do. Write it down as a safety policy before you build.

  2. Align

    Shape the model's behaviour to that policy, through prompts, examples or tuning, and look into why it responds as it does.

  3. Evaluate

    Test against your policy and compare versions side by side, so you know whether a change helped.

  4. Protect

    Add checks around the model: classify what comes in and what goes out, and label generated content.

Stage names from Google's Responsible Generative AI Toolkit; the one-line explanations are this site's summary. Google also points to its Secure AI Framework for security.

What is available

Safety tools you can add

ToolWhat it doesDetails
ShieldGemmaChecks text going in or coming out against a safety policy and answers yes or no2B, 9B and 27B sizes, built on Gemma 2. Covers hate speech, harassment, dangerous content and sexually explicit information.
ShieldGemma 2Checks images against a safety policy4B, built on Gemma 3. Covers sexually explicit, dangerous and violent content.
SynthID TextWatermarks generated text so it can be identified laterPart of Google's toolkit
Agile classifiersA method for building your own safety classifier from a small number of examplesPart of Google's toolkit
LLM ComparatorRuns and visualises side-by-side evaluation resultsPart of Google's toolkit
Learning Interpretability Tool (LIT)Helps investigate and debug how the model responds to promptsPart of Google's toolkit
Two things to know about ShieldGemma. It is released under Google's older Gemma Terms of Use, not Apache 2.0. And Google's own cards say it is highly sensitive to how you word the safety policy, and may be unpredictable where language is ambiguous. Treat it as one layer, not a guarantee.

From the ShieldGemma and ShieldGemma 2 model cards and the Responsible Generative AI Toolkit. Neither ShieldGemma version is built on Gemma 4.

EmbeddingGemma 2 · released October 6, 2026
The search model

EmbeddingGemma 2: ethics and safety

EmbeddingGemma 2 does not write text. It turns text, images, audio and video into numbers used for search and sorting. That changes the safety picture: there is no harmful answer to block, but the model can still surface the wrong thing, or rank people unfairly.

EmbeddingGemma 2 finds; your filter decides what is shown Search queryturned into a vector Annual report 2025shownTeam handbookshownLeaked customer listfiltered outProduct FAQshown The result filter is something you build. The model does not include one.
Safety here means controlling what your search returns.
⚠️

No safety tuning

Google's card says the model has no post-training alignment or safety tuning. Safety work was focused on filtering the training data.

🧹

Same data filtering

Child-safety filtering at multiple stages, removal of certain personal and sensitive data, and quality and safety filtering.

📜

Licence and policy

Apache 2.0. The model card also says deployments must follow the Gemma Prohibited Use Policy.

Risks Google names

  • Bias. The vectors can reflect social and cultural biases in the training material, which can skew search results or classifications.
  • Misuse. Embeddings can be used to retrieve, classify or organise content in false, misleading or harmful ways.
  • Privacy. Developers are told to follow privacy regulations and use privacy-preserving techniques.
  • Uneven languages. Performance varies across the 100+ languages it supports.

What that leaves to you

  • Filter results. Google names retrieval filtering as an application-level safeguard that is your responsibility.
  • Test for fairness, especially if results affect people: hiring, lending, moderation.
  • Control what you index. A search can only surface what you put in it. Keep private material out, or behind access rules.
  • Treat stored vectors as sensitive. They are derived from your documents, so protect them as you would the originals.

The left column and the first two points on the right are from Google's EmbeddingGemma 2 model card; the last two are this site's guidance. Running it privately is covered in how to use and the prompt guide.

Practical

Deployment checklist

A minimum bar for putting Gemma 4 in front of other people. Tick items off as you go; nothing is saved or sent.

0 of 13 done

Before launch

After launch

Extra care in high-stakes areas. For medical, legal, financial or similar uses, do not let the model make decisions on its own. Google's policy prohibits automated decisions in high-stakes domains and the unlicensed practice of professions; keep a qualified person responsible.

The checklist is this site's guidance, drawn from Google's toolkit stages. It is a starting point, not a compliance standard. Personal use on your own machine needs far less; see chat and the playground.

FAQ

Common questions

More answers on the full FAQ page.

Is Gemma 4 safe to use?

Google filtered its training data and reports that it outperforms Gemma 3 on safety in testing. No model is safe in every situation, and a downloaded model has no extra filters around it unless you add them.

Does Gemma 4 have built-in content filters?

The model was trained to avoid harmful output, and Google tested it with filters off. There is no separate filter in the download. Hosted services may add their own.

Is it uncensored?

No. It is trained to decline harmful requests. Because the weights are open, others can modify it; a modified version is not the model Google tested.

What does Apache 2.0 mean for safety rules?

The licence itself sets no limits on use. Google publishes a Prohibited Use Policy and an Intended Use Statement alongside it. For how these apply to you, take legal advice.

Can I use ShieldGemma with Gemma 4?

Yes, as a separate checking step. It is a different model with its own terms, built on earlier Gemma versions.

Is running it locally more private?

Yes. Prompts and files stay on your machine. You remain responsible for how you use the output.

Does fine-tuning affect safety?

It can. Tuning changes the model's behaviour, so re-run your safety tests afterwards.

Is EmbeddingGemma 2 safe?

It cannot generate harmful text, but it has no safety tuning. What your search returns, and whether it treats people fairly, is for you to check.

Know the limits before you build

Safety starts with knowing what the model can and cannot do.

Top