Global: AI - A Single Global Approach | iGlobal Law

While AI races forward, regulation is in the slow lane. But like the tortoise, it will at last get somewhere. Although, like Zeno’s Achilles, it won’t ever overtake.

Different regulatory approaches are already visible in different regions/jurisdictions. A question for multi-country employers is whether to adopt bespoke local AI workplaces policies, to satisfy each set of local regulations, or one policy that likely meets all regulations.

If the one global policy approach is at all possible, the starting point would be to understand the workplace risks that AI clearly presents, as these will no doubt become in due course the subject of local regulation.

Generic AI Workplace Risk Areas

Employer/employee mutual trust and confidence – essential to productivity/talent retention.
Equality and fairness – hidden (black box) algorithmic discrimination/bias (e.g. older workers maybe more harmed than younger, lower skilled more than higher; female more than male; disabled more than abled).
Privacy (AI analysis of sensitive personal information – e.g. bio-metric data, personality type etc.).
Data protection and security – not new but more vulnerable.
Health and Safety – mental stress related.

Specific AI Workplace Usage Risks

Recruitment:

Sourcing (e.g. source pre-selecting, job describing, chatbots used to guide applicants)
Screening (e.g. volume application sifting – but on what algorithm)
Assessment and psychometric testing (e.g. online testing of analytical skills, personality mapping)
Interviewing (e.g. assessing online interviews by bio-metric data – speech patterns, tone, facial movements, AI analysis of non-verbal communication)

Line Management:

Shift scheduling (e.g. just in time work allocation)
Performance (e.g. assessment of performance/productivity, promotion assessment, data for redundancy selection)
Monitoring and Surveillance (e.g. biometric monitoring, face recognition, skills assessment, productivity assessments, real time fatigue or stress assessments)

Common Themes of Emerging Regulation

Regulation approaches visible so far:

Risk based formal approach – European Union AI Act
Best practice guideline approach – US Eight Principles and Best Practices Executive order (Federal)
Individual regulations – China (existing); UK (indicated but not yet published); Individual US States (e.g. Colorado – effective 2026)

Developing a Global AI Policy

If a global employer should choose to develop a global approach, rather than a patch work of local policies, in response to real or anticipated local or regional regulation, the policy should cover the following ground:

Issue	Response
Risk Assessment	Employers should carry out and regularly review an informed AI risk assessment for each AI tool used. The EU AI Act states that AI systems used for social scoring that can cause harm are prohibited. It further perceives workplace AI use as generally “high risk” – to which specific rules apply, including registration.
Employee Involvement	AI will affect all businesses and everyone (owners, managers and staff) who work in them. It is essential to the trust and confidence relationship that employers work closely and openly with employees on AI developments and use, so that the benefits are understood and accepted and the consequences (e.g. work changes and job redundancies) are well managed.
Human Oversight and Governance	No decisions affecting people should be taken without a safe level human oversight.
Transparency	Given the Black box risk and the extensive use of machine learning, AI systems need to be kept transparent and explainable so that users and effected individuals can understand the reasoning behind decisions.
Health and Safety	Existing employer obligations to provide a safe workplace environment will apply to AI. Given the possible risks to mental health of AI applications, appropriate safeguards should be put in place.
Rights Protection	Acknowledgement that all employee legal rights are protected and not undermined by AI applications.
Support for Change	Recognition that the AI revolution will have significant consequences for life styles and livelihoods. Where jobs will be lost or materially changed, put in appropriate support, such as re-training options.
Responsibility	Employers should accept that they have a responsibility for the risks that AI might give rise to. There should be clear lines of responsibility and accountability within organisations. This includes mechanisms to challenge and rectify AI based decisions that have adverse or unintended effects.

This is a high level general update only. Legal advice should be obtained on specific circumstances.