Cross-Border: AI Driven Redundancies

Sweeping multi-jurisdictional worker redundancies have been announce recently by Microsoft, Intel, Google, Amazon and IBM. Replacement of workforce by AI will have been a factor in all of them. 

What are the risks associated with mishandling multiple AI driven layoffs across borders? 

While a global redundancy programme may well be lead from the business centre, the risks and therefore the best approach to take, will be local (i.e. by jurisdiction).

The costs of getting it wrong locally, can be very large.

Valid Reason

Most jurisdictions permit dismissal for fair business reasons (such as a reduced requirement for a particular task or in times of business downturns).  However, employers are usually required to show that the business reason is valid.  In the case of largescale AI redundancies, employers may find local courts applying greater levels of scrutiny and challenge to the question of what a fair business reason is.

Collective Redundancies

Most countries have rules governing mass (or collective) redundancies.  These usually trigger an obligation to carry out formal consultation, sometimes involving works councils, trade unions or other forms of worker representatives. Some jurisdictions require prior notification of the redundancies to state labour authorities or even the courts.  Errors in notifications and consultation can significantly raise the cost of an individual dismissal.

Notice Periods

There will be both contractual and statutory notice periods to respect.

Obligations to Consider Alternatives to Redundancy

Some jurisdictions require employers to attempt to find suitable alternative roles before making an individual redundant. Failure to consider properly this redeployment option can result in expensive unfair dismissal claims, even where no such option existed.          

Fair Selection Process

The pooling of employees for consideration for redundancy must be reasonable and objective and aligned to local rules. Great care needs to be taken to ensure the selection process does not inadvertently, disproportionately affect protected characteristics (such as age, race, gender, religion, sexual orientation or disability). For example, it may prove dangerous to use “last in, first out” selection (more likely to impact younger employees) or requiring a degree level education (which may impact older employees).

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