Skip to main content ->
Ai2

Responsible use guidelines

Last Updated: June, 2024

Introduction:

These Responsible Use Guidelines (“Guidelines”) govern your use of Ai2’s AI tools and artifacts, including models, datasets, APIs, and applications (the “Ai2 Tools”). In addition to these Guidelines, your use of the Ai2 Tools is governed by Ai2’s Website Terms and Conditions of Use and Privacy Policy. As used in these Guidelines, “you” refers to any user of the Ai2 Tools, and “we”, “us”, or “Ai2” refers to The Allen Institute for Artificial Intelligence, a 501(c)(3) nonprofit organization.

Our Goals:

The Ai2 Tools are intended for research and educational purposes and are provided on a free and open basis for the benefit of the general public. More specifically, the Ai2 Tools are intended to advance fundamental scientific research about AI and to empower academics and researchers working on breakthroughs in the fields of AI and machine learning. By opening up the AI Tools, our goal is to encourage innovation and address the technology gaps that prevent us from fully understanding AI.

Responsible Use Principles:

We are mindful of safety concerns relative to the variety of audiences this technology impacts. As a result, some of our Ai2 Tools may need to be adapted for more limited use and exploration by the general public, whereas AI researchers may be allowed broader access to research technology gaps that underpin AI safety issues. You should understand the Ai2 Tools and their limitations, which will change as the Ai2 Tools improve. Please also refer to the applicable model cards or dataset cards for more information about the artifacts.

Good Judgment Required:

Please use your own judgment to review and validate generated outputs from the Ai2 Tools: no outputs should be accepted at face value and users are expected to evaluate all outputs critically. Outputs generated by the Ai2 Tools may be inaccurate, misleading, or otherwise reflect content that does not represent the views of Ai2. Never rely on outputs from the Ai2 Tools without conducting your own independent consultation or research. The Ai2 Tools are not intended to, and cannot, replace advice provided by qualified professionals; your use of the Ai2 Tools does not form any qualified professional relationship of any kind between you and Ai2.

Understand and Prevent Disallowed Use Cases:

These Guidelines do not restrictthe use of the Ai2 Tools for the sole purpose of scientific research and development (“Research Use”). Except for Research Use, the Ai2 Tools may not be used for any of the disallowed use cases below. The descriptions are intended to be illustrative and not exhaustive. Ai2 reserves the right to terminate access for harms that are not listed at our sole discretion.

1. Violence and threats:

a. Violence/Incitement: Actions that threaten, encourage, or incite violence against anyone, directly or indirectly.

b. Self-harm: Promoting or glorifying acts of self-harm, such as cutting, eating disorders like anorexia or bulimia, and suicide.

c. Sexual exploitation: Promoting or celebrating sexual exploitation, including the sexualization of minors.

d. Hate speech: Promoting hatred or glorifying abuse against people based on characteristics like race, ethnicity, national origin, religion, disability, disease, age, sexual orientation, gender, or gender identity.

2. Antisocial and antidemocratic uses:

a. Harassment: Bullying, threatening, shaming, or doxing.

b. Insensitivity: Belittling victims of serious physical or emotional harm (even if unintentional).

c. Intentional sowing of division: Sharing of divisive generated content in order to turn a community against itself.

d. Harmful belief perpetuation: Perpetuating racism, or sexism (even if unintentional).

e. Applications that aim to characterize identity: Attempting to characterize gender, race, or ethnicity.

f. Graphic depictions: Distribution of sexually explicit acts, torture, or abuse.

g. Political manipulation: Attempting to influence political decisions or opinions.

3. Deceit:

a. Fraud: Catfishing, phishing, or attempting to circumvent the law.

b. Spam: Sending unsolicited emails and messages, or manipulating search engines.

c. Misrepresentation: Representing raw generations as coming from humans, using supervised generations with false identities, or a single person using generations with many identities that appear to be independent.

d. Misinformation: Creating or promoting harmful false claims about government policies, or public figures, including applications founded on unscientific premises.

4. Attacks on security or privacy:

a. Security breaches: Spearphishing.

b. Malicious code: Creating malicious code, malware, computer viruses or any activity that could disable, overburden, interfere with or impair the proper working, integrity, operation or appearance of a website or computer system.

c. Privacy violations: Model attacks to extract personal information.

5. Unsafe unsupervised uses:

a. Social media: Posting content to social platforms in an automated way.

b. No transparency: Applications that do not disclose that the content is generated through automated means.

6. Decision-making:

a. AI-based social scoring for general purposes done by public authorities; using output toward larger decision-making systems that will influence actions, decisions, or policies without a human in the loop.

b. Classification of individuals: Applications that classify and/or profile people based on protected characteristics, or infer those characteristics from text written about them or by them.

7. Other:

a. Intentional manipulative redirection of attention; sharing positive generated content in order to direct attention away from harmful actions.

b. Plagiarism: Tools that promote academic dishonesty.

c. Political campaigning or lobbying.

Changes to these Guidelines

We reserve the right to update or modify these Guidelines at any time. Users are encouraged to review these Guidelines periodically to stay informed of any changes. Any such changes will not apply retroactively to any Ai2 Tools that you may have already accessed or used.

Questions & Reporting

If you have any questions about these Guidelines or need to report violations, please contact us at support@allenai.org. Please report any “bugs” in the AI Tools, security concerns, violations of these Guidelines, or any other problems through one of the following means: