Assess the risks of AI use
across your teams
Shadow AI, hallucinations, safety… Taboo puts your employees in interactive work scenarios to assess their vulnerability to the main risks of AI.
10 min
interactive test
5
types of risk
100 %
anonymous
No
installation required
Taboo assesses 5 risk types in real work situations
Give me a figure on AI adoption in customer service.
According to a 2024 Gartner study, 78% of customers prefer an AI agent to a human.
Memo to the executive committee
Finally know where your teams stand
A 10-minute interactive test based on real work situations, with anonymous results benchmarked against your whole company and your industry.
1Identify the risks within each team
Marketing · Overall score
58/100
- vs industry
- +4 pts
- Priority
- Safety
2Train the teams that need it most first
- 1Sales6,400 employees41
- 2Customer support3,100 employees52
- 3Marketing4,200 employees58
- 4Human resources1,200 employees63
- 5Finance1,800 employees68
Sample data, for illustration
AI comes with risks to manage
Legal risk
Confidential data entered into unapproved tools puts the company's liability at stake (GDPR, AI Act, trade secrets, etc.).
Reputational risk
Hallucinations and leaked information undermine the trust your partners and clients place in you for a long time.
Operational risk
Decisions based on unverified answers undermine the quality of work and your teams' productivity.
Compliance · AI Act
Comply with the AI Act
Since February 2025, the AI Act has required companies to ensure their employees are AI literate and use AI responsibly.
- Assess with Taboo
- Train with Ritual
- Document progress
Assess your organisation's AI risks today
Frequently asked questions
How long does the test take?
The test takes around 10 minutes per employee, on their own computer. No installation is required, and everyone can take it whenever suits them.
Are the results really anonymous?
Yes. Results are only ever presented in aggregate, by team or business unit. No personal data appears in the final report. Taboo assesses risks at organisational level, not individual performance.
Is any installation required?
No. The test runs directly in the browser. It requires no installation, no account creation and no involvement from your IT department.
What happens after the test?
You receive a detailed report: a score for each risk, team and business unit, benchmarked against your company and industry averages. You can then roll out Ritual to the teams most at risk.
Already tested your teams?
Switch to
Once Taboo has identified the risks, Ritual trains the relevant teams in 1 hour a year: 10 video modules updated every year, a final test and a certification backed by an industry consortium and listed by the European Commission. It plugs straight into your LMS or runs on our own platform.
10
video modules
1 h
per person per year
Certification
backed by an industry consortium
LMS
direct integration (SCORM)
Use AI safely and responsibly
in 1 hour a year
Short and practical modules, built around real use cases and updated every year.
Why train with Ritual
An all-in-one certified programme for all types of profiles,
to stay up to date on risks and solutions.
Meet the AI Act's training requirements and align your teams with international standards to empower them to boost productivity with AI, responsibly and safely.
Protect yourself from operational and reputational risks: train your teams to identify and resolve issues with best practices.
Bring your teams on board with a shared vision, built on a common vocabulary and a single information channel, to keep them up to date with the latest innovations.
Designed by recognised experts
Ritual is designed and presented by scientific experts specialising in AI ethics and safety.
Dr Hubert EtienneCEO · Quintessence AI
Ex-Global Head of GenAI Ethics at Meta. Associate Professor at Sciences Po and HEC Paris. Visiting researcher at Oxford, Stanford and Harvard. PhD from ENS Ulm. Member of the French Ministry of Justice's AI Observatory.
Dr Jean-Marie John-MathewsCEO · Giskard AI
AI safety researcher. Former data scientist at Thales and holder of the Good in Tech Chair (Sciences Po/IMT). Graduate of École Polytechnique, ENSAE and Paris-Saclay.
Dr Giada PistilliModel Behaviour Architect · Mistral AI
Former Principal AI Ethicist at Hugging Face. Associate researcher at the Sciences, Norms & Democracy Lab (Sorbonne/CNRS). PhD in AI ethics from Sorbonne University.
Consortium
Training backed by a leading industry consortium
The Ritual certification is backed by a consortium of industry leaders and public actors. It sets a robust market standard to support our ambition of training one million professionals.
AI will only create lasting value at Safran if it rests on a foundation of responsibility and trust. By supporting Quintessence AI, we are choosing a concrete, annual programme to build our teams' understanding of the ethical use of AI.
The great challenge of the coming years is not technological: it is human and collective.
Grounded in science
Ritual is designed and evaluated by the Quintessence AI research team, together with our consortium. Content is updated every year to keep pace with changing tools and uses, risks and standards, scientific discoveries and best practices.
- Teaching for every audience
- Designed by world-class scientists
- Updates driven by ongoing research
A certification listed by the European Commission
Each employee receives a personal certificate on completing the training. It attests to a solid grounding in the responsible use of AI and demonstrates that your teams have been trained.
- Meets the AI Act's training obligations
- A personal certificate for every employee
- Renewed every year with the latest version
AI comes with risks to manage
Legal risk
Confidential data entered into unapproved tools puts the company's liability at stake (GDPR, AI Act, trade secrets, etc.).
Reputational risk
Hallucinations and leaked information undermine the trust your partners and clients place in you for a long time.
Operational risk
Decisions based on unverified answers undermine the quality of work and your teams' productivity.
Compliance · AI Act
Designed to meet Article 4 of the AI Act
Since February 2025, the AI Act has required companies to ensure their employees are AI literate and use AI responsibly.
- Assess with Taboo
- Train with Ritual
- Document progress
Train your teams today
Frequently asked questions
How long is the training?
The training takes around one hour per employee per year: 10 short video modules followed by a final test. Everyone progresses at their own pace and picks up where they left off.
Can Ritual be integrated into our LMS?
Yes. Ritual integrates directly into your learning platform, so you can track progress with your existing tools.
Where is the data hosted?
Data is hosted in Europe and processed in compliance with the GDPR. Only the information strictly required to track the training is retained, solely for that purpose.
Have you already trained your teams?
Test them with
In 10 minutes, Taboo assesses what your teams have really retained,
risk by risk, in real situations.
They trust us
- Equimetrix
