G7 and AI Companies: How New Rules for Safe AI Use Will Impact Business Management

What the Unprecedented G7-AI Executive Lunch Meeting Signals

In April 2025, an unprecedented lunch meeting brought together G7 leaders and top executives from global AI companies. The theme was “safe use of advanced AI.” Participants included OpenAI CEO Sam Altman, Google CEO Sundar Pichai, and Anthropic CEO Dario Amodei, who exchanged views on an international AI governance framework.

This news is more than just a diplomatic event. For business leaders, it signals that the “rules” for AI use are about to change significantly. If regulations tighten, the use of AI tools that were previously unrestricted could face limitations. Conversely, companies that meet safety standards will find new business opportunities.

Why Safe AI Use Now? The Background Business Leaders Need to Know

AI evolution is accelerating at an exponential rate. Particularly from late 2024 to 2025, the performance of generative AI has improved dramatically. The market is becoming multipolar, with ChatGPT’s global share falling below 50% while Gemini and Claude are surging.

Behind this shift lies the fact that corporate AI use has moved from the “experimental stage” to the “operational stage.” If AI integrated into business operations makes a wrong decision, it can cause direct damage. For example, if an AI-generated contract contains legal errors, it could directly lead to litigation risk.

The G7’s urgency around safe use stems from these risks becoming a reality. What business leaders cannot ignore is that regulatory movements are shifting from “self-regulation” toward “legally binding rules.”

Three Management Risks of Safe AI Use

Legal Risk: Unclear Accountability

If a problem arises with AI-generated content, who is responsible? Currently, in most cases, liability falls on the company that introduced the AI. In Japan, lawsuits related to copyright and privacy violations from AI-generated content are increasing.

Specific examples include cases where AI mistakenly leaked customer data externally, or where generated text infringed on another company’s copyright. These risks could significantly increase the cost of AI implementation.

Compliance Risk: Regulatory Uncertainty

The EU’s AI Act will be phased in starting in 2025. In Japan, guidelines on AI are also expected to be strengthened. AI tools currently used freely could suddenly become “regulated.”

Particularly, the use of AI for facial recognition, biometric data processing, and employment decisions will face strict regulations. Business leaders need to quickly determine which category their company’s AI use falls into.

Brand Risk: AI Malfunctions Damage Trust

Cases where AI malfunctions or biased outputs severely damage a company’s brand image are increasing. For example, an incident where a customer support AI gave an inappropriate response spread on social media, significantly eroding the company’s credibility.

This risk exists for any company using AI. It can be particularly devastating for B2C businesses.

Three Actions Business Leaders Should Take Immediately

Establish AI Governance

First, formalize internal rules for AI use. Specifically, define the following:

– Scope of AI use and prohibited actions
– Review process for AI-generated content
– Accountability for AI-related issues
– Regular auditing and improvement mechanisms

For my client companies, building AI governance typically takes about 2-3 months. Implementation costs are roughly several hundred thousand yen (approx. $3,000-$7,000) using internal resources, or 1-3 million yen (approx. $7,000-$20,000) if hiring an external consultant.

Conduct Regular AI Audits

A system for regularly auditing AI outputs is necessary. The following three methods are effective:

1. Manual sampling checks
2. Introduction of AI audit tools (e.g., Credo AI, Monitaur)
3. External audits by third-party organizations

Monthly audits are recommended, especially when AI handles customer data or confidential information. Audit costs range from 100,000-500,000 yen per month (approx. $700-$3,500) for tool implementation, or 500,000-1 million yen (approx. $3,500-$7,000) per session for external outsourcing.

Implement AI Literacy Training

To use AI safely, improving the literacy of every employee is essential. From management to the front line, all employees need to understand AI risks and proper usage.

Training should include the following:

– Basic AI mechanisms and limitations
– Fundamentals of privacy and security
– How to verify AI-generated content
– Incident reporting procedures

Training costs are minimal (tens of thousands of yen, or a few hundred dollars) for in-house materials, or 200,000-500,000 yen (approx. $1,400-$3,500) per session if hiring an external instructor.

The New Trend: FDE-Type Support Service “Forward Runner”

Alongside these developments, a noteworthy service has emerged: “Forward Runner,” an FDE-type support service that guides companies to in-house AI and RPA implementation.

The key feature of this service is that it aims for in-house implementation as the ultimate goal, not just tool introduction support. To achieve safe AI use, a system that the company itself can control, rather than relying on external parties, is essential.

Services like Forward Runner are effective options for companies aiming for long-term in-house implementation while keeping initial AI introduction costs down. Monthly fees are estimated at 500,000-2 million yen (approx. $3,500-$14,000), depending on the service content.

The Future of Safe AI Use and the Business Leader’s Choice

The G7-AI executive meeting shows that safe AI use is shifting from an “option” to a “requirement.” Business leaders have three choices:

1. Wait for regulations before acting (Risk: Falling behind)
2. Voluntarily implement safety measures (Recommended)
3. Accelerate AI use while simultaneously implementing safety measures (Ideal)

From my experience, safety measures and AI use are not a trade-off. Rather, safe AI use creates long-term competitive advantages by earning trust from customers and partners.

As a concrete first step, start by taking stock of your company’s AI usage. Visualize which tasks use AI and what risks may be lurking. Then, prioritize and proceed with safety measures.

Used correctly, AI can be a powerful management tool. But used carelessly, it can endanger your company. View the G7’s moves as an opportunity and build a foundation for safe AI use now.