Alt-TAB Assessment <span aria-hidden="true">—</span> Think. Assess. Build Ethically.

Check your product for risks before you launch.

Free, no account needed, and around 15 to 20 minutes. Built for community groups, sole traders, NGOs, and small businesses as much as for developers and tech teams. Covers products, services, processes, and policies.

"Unsure" is a valid response and often the most useful one. There are no wrong answers.

Nothing is saved between sessions. You can close this tab at any time.

You will receive a full report: a readiness score, gaps identified, a 30-day action plan, and questions for your team.

Questions?

Do I need to be technical?

Not at all. Plain language throughout. Designed for community groups and sole traders just as much as developers.

What if I don't know an answer?

"Unsure" is a valid answer. The tool is built to surface what you haven't thought about yet, not to test what you know.

What do I get at the end?

A readiness score, a breakdown of gaps found, specific recommendations, a 30-day action plan, and questions to take to your team. Downloadable as a PDF.

Is this legal advice?

No. It identifies legal obligations and which law applies, but it's a decision-support tool, not legal advice.

Domain questions

Deeper dive

Your report

Question 1 of 11

Just getting started

Domain 1 of 9High sensitivity

Could children or young people interact with, or be affected by, this product?

Consider direct users, indirect exposure through family members, or products deployed in schools or community settings. "Maybe" means the risk exists but you're uncertain of the extent. "Unsure" means you genuinely haven't considered whether this applies.

We ask this in every assessment. Children encounter technology designed for adults far more often than builders expect, through family accounts, community platforms, and apps with no age restriction.

Question 1 of 11

Domain 2 of 9High sensitivity

Could your product affect people who are vulnerable, marginalised, or at greater risk of harm, such as people experiencing family violence, disability, poverty, or discrimination?

Think beyond your intended users. Who else might encounter, be affected by, or be excluded from this product?

Products built for "general users" regularly reach people in crisis, people with disability, First Nations communities, or people experiencing family violence. Unintended reach is one of the most common gaps we find.

Question 2 of 11

Domain 3 of 9High sensitivity

Does this product collect, store or use data that could be accessed or reused in ways users may not reasonably expect or consent to?

Include biometric data, location, behavioural patterns, health information, and financial details. Also consider: any data your product shares with third-party tools like Google Analytics, Facebook Pixel, or cloud providers, even if you don't think of it as "collecting" data, those services may be.

Data practices are the most litigated area of technology law in Australia. Most builders underestimate how much data their product collects and how little protection it has.

Question 3 of 11

Domain 4 of 9Flagged

Does this product enable private interactions, anonymity or content sharing, directly or through third parties, platforms or integrations?

Include messaging features, anonymous accounts, user-generated content, or connections to external platforms. Also consider: does your product connect to any third-party services, analytics tools, payment processors, AI APIs, cloud storage, or social media integrations? Those connections can create exposure you haven't designed for.

Private interaction features, messaging, anonymity, shared accounts, location sharing, are the most common pathway for technology-facilitated gender-based violence. This is screened in every assessment.

Question 4 of 11

Domain 5 of 9Flagged

If something went wrong with your product, could this cause serious harm to individuals: emotional, reputational, sexual or financial?

Consider worst-case scenarios: a data breach, a system error, misuse by a bad actor, or harm caused by an automated decision.

The severity of potential harm shapes every recommendation in your report. Products that could cause serious harm require a higher standard of care before deployment.

Question 5 of 11

Domain 6 of 9Flagged

Do you have documented safeguards in place to prevent, detect, or respond to foreseeable harm?

Include content moderation, incident response plans, abuse reporting tools, security controls, or human oversight of automated decisions.

Your description will be cross-referenced with this answer. Be specific about what safeguards you actually have in place. vague answers here are less useful than honest ones.

Question 6 of 11

Domain 7 of 9Flagged

Is there a named person or team responsible for safety outcomes if something goes wrong with this product?

This includes accountability for data breaches, harm to users, regulatory compliance, and decisions about when to pause or withdraw the product.

In 37 assessments, unnamed accountability has been one of the most consistent gaps. When something goes wrong, someone needs to be responsible. That person needs to be named before launch.

Question 7 of 11

Domain 8 of 9Lower sensitivity

Would users understand how this product affects them, including how their data or interactions are used across different systems?

Consider whether your privacy policy, terms of service, and in-product explanations are written in plain language a non-expert could genuinely understand.

Users who don't understand what a product does with their data can't give meaningful consent. This is a legal obligation under Australian privacy law, not just good practice.

Question 8 of 11

Domain 9 of 9High sensitivity

Is your product designed to be accessible and usable by people across different abilities, languages, cultural backgrounds, and socioeconomic circumstances?

Consider WCAG accessibility standards, multilingual support, low-bandwidth performance, and whether First Nations or CALD communities have been consulted.

One in five Australians lives with disability. Inaccessible technology is discriminatory under the Disability Discrimination Act 1992. WCAG 2.2 Level AA is the compliance standard.

Your descriptionAI analysis

Now describe what you are assessing in your own words: what it does, who it affects, what data it involves, and any features where people interact or share information.

Keep going. Aim for at least 4 to 5 sentences.

0 words

You've answered the domain questions — now describe your product as it actually is. Your description is compared against your answers, so write what is genuinely true. Aim for 48 sentences and be specific about features, users, and data.

Question 10 of 10

Everything you type exists only in your browser. Alt-TAB processes it in real time and discards it immediately. How this works

Stage 1 complete

Reviewing your answers...

The questions below are written specifically for what you described. This is where the score is shaped.

Stage 2 of 2

Let's go deeper

Answer honestly. Your readiness score weights these answers heavily, a thoughtful "partially" with explanation scores better than a confident "yes" that doesn't hold up.

How these answers shape your score

YesFull credit, you have addressed this

PartiallyHalf credit, detail in the text box improves this

NoNo credit, but honesty produces better recommendations

These questions carry 50% of your readiness score. Honest, specific answers are more useful than optimistic ones.

Tell us about your deployment strategy and any safeguards already in place (optional, but this shapes 20% of your readiness score)

The more detail you provide here, the more accurately your readiness score will reflect your actual preparedness.

Your Ethics & Safety Assessment Report

This report was generated by Alt-TAB, built by Away from Keyboard Inc. It checks your situation against 35 Australian and international governance frameworks and is a structured decision-support tool, not legal or compliance advice. Completing this assessment does not certify your product as safe or compliant. Use it as a starting point for deeper investigation. Not sure what a framework is? Every one is explained at alttab.afk.org.au/resources

Start a new assessment

Was this assessment useful?

How your privacy is protected

What was never stored: your product description, your answers to the deeper dive questions, your deployment strategy text, your IP address, and anything that could identify you or your product.

What is stored if you gave consent: anonymised aggregate counts only: your yes/no/maybe/unsure answers, your readiness score band, and your awareness level. These are added to a running total with no link to you. No individual record exists.

How Alt-TAB works: your description and answers were sent to Anthropic's API (which powers Alt-TAB) and discarded immediately after this report was generated. Anthropic does not use API inputs to train its models by default. Away from Keyboard's server never wrote your description to disk and cannot access it. Anthropic's privacy policy · github.com/afk-org/alt-tab, or by getting in touch at info@afk.org.au.

Alt-TAB is currently seeking independent academic peer review of its assessment methodology. This report is a decision-support tool and should be used alongside, not in place of, specialist advice where serious risks are identified. About our limitations

Assessment record

Reference

Generating...

Date

Score

Type

This reference code confirms an assessment was completed on this date with this score. It does not contain your product description, your answers, or anything that identifies you or what you assessed. Share it as evidence of due diligence with a board, funder, or regulator. How verification works · Request a co-signed record

Run a different product through Alt-TAB, or retake this one after addressing the gaps

Want to go further?

AFK can work directly with organisations that want to close the gaps identified in this report. Get in touch at info@afk.org.au.

Get in touch: info@afk.org.au

Did this assessment help you think differently? Share your experience

Ethics & Safety Assessment

Check your product for risks before you launch.

What are you working on?

Let's go deeper

Your Ethics & Safety Assessment Report