RSM France integrates "Think First, Verify Always" protocol for responsible AI usage

RSM France, the 6th largest global network in audit, accounting, and consulting, has integrated the "Think First, Verify Always" protocol for responsible AI usage.

This protocol was developed and deployed by Yuksel Aydin, CISO and DPO of RSM France, as part of an initiative to integrate AI into RSM processes.

The TFVA protocol is an operational method designed to strengthen human vigilance when interacting with artificial intelligence tools.

As part of its strategy for controlled AI integration, RSM France has adopted this protocol, enabling all employees to apply it in their daily practices.

Paris, December 8, 2025

Yuksel Aydin, Chief Information Security Officer (CISO) and Data Protection Officer (DPO) of RSM France, as well as an independant researcher, has developed and deployed the "Think First, Verify Always" (TFVA) protocol - an operational method intended to enhance human vigilance when using artificial intelligence tools. This protocol was created within the framework of an initiative to integrate AI into RSM processes.

With the rapid evolution of AI technologies, methods of manipulation and deception targeting human cognition are becoming increasingly complex.

In response to this threat, Yuksel Aydin created the TFVA protocol, which is human-centric and positions the user as the first line of defense against AI-generated misinformation.

The protocol stems from a key observation: traditional cybersecurity strategies, focused on systems and technical safeguards, are no longer sufficient. Attacks now directly target users through phishing, social engineering, or artificially generated content. As a result, the ability to think independently and verify information is essential.

The "Think First, Verify Always" protocol

Think First, Verify Always is based on two concrete principles :

• Think First : users are encouraged to engage in independent reasoning before accepting or acting on information generated by AI. This step helps limit cognitive biases and avoid automatically generated responses that could be misleading.
• Verify Always: encourages the systematic verification of results provided by AI, particularly when they have operational or decision-making impact. This may include consulting reliable sources, cross-checking data, or seeking human expertise to validate information. This step ensures that AI remains a decision-support tool rather than an exclusive source of trust.

These two steps are structured around five operational principles, grouped under the acronym AIJET : 

• Awareness ;
• Integrity ;
• Judgment ;
• Ethical Responsability ;
• Transparency.

Demonstrated effectiveness

The effectiveness of the protocol has been proven: a randomized controlled trial involving 151 participants showed that an intervention of just 3 minutes based on this protocol significantly improved cognitive performance in the face of digital risks. Trained participants demonstrated a 7.87% improvement compared with the placebo group.

This level of effectiveness is remarkable considering the protocol's speed and simplicity. Unlike traditional security training - often long and complex - TFVA provides concrete behavioural guidelines that are easy to remember and apply immediately.

Commited to the controlled adoption of AI, RSM has now integrated the "Think First, Verify Always" protocol into its internal training and presentations. Employees learn to apply these two steps in their daily professional activities, thereby strengthening work quality and building trust in the use of AI. The method is designed to be immediately applicable et adaptable to all professions and experience levels.