Indian boards that blithely embrace fads should weigh the legal liability risks of reckless AI adoption

Insurers dislike uncertainty with the same passion Indian parents reserve for unconventional career choices. AI, unfortunately, is uncertainty at an industrial scale. Nobody fully knows where liability begins, where accountability ends or who exactly would be responsible for an algorithm that behaves as if it was schooled in ethics by an anonymous online forum. So insurers are responding the only way they know: with exclusions.

Some of these are subtle. Some are broad. Some essentially say: ‘If AI was anywhere near this catastrophe, don’t call us.’ Globally, legal experts have warned that insurers are inserting AI-related carve-outs into D&O policies. Since most boards read insurance documents like they examine software terms and conditions, many directors may not notice until there’s a regulatory run-in.

Imagine the possibilities. A company exaggerates its AI capabilities to investors. A chatbot leaks customer data. An automated hiring tool begins to reject applicants with particular names. A financial forecasting model hallucinates growth projections with the confidence of a startup founder. Shareholders would be right to ask difficult questions. Regulators ought to be looking in. And D&O insurance coverage wouldn’t help.

Across sectors, AI adoption is accelerating faster than governance frameworks around it. Companies want the valuation premium of appearing AI-driven, even if safe implementation is a nightmare. This sets the stage for what lawyers call ‘AI washing’—the corporate art of taking ordinary automation, sprinkling the letters A and I over it and presenting it as a productivity leap.

Investors may be told that advanced AI-powered customer engagement ecosystems are in place, while all that has happened is the deployment of a chatbot adept at adding to the frustration of complainers. What happens when the emperor’s algorithm is found to have no clothes?

Lawsuits linked to exaggerated AI claims have begun to emerge, globally. Directors signing off on disclosures may find themselves personally exposed if oversight appears weak. It is thus clear that boards must adopt AI with care. In some cases, it is like radioactive material: useful and transformative, but in need of safety protocols.

What should companies do?: First, conduct an AI inventory. Map all AI use across the organization. Which tools? Which departments? What data enters these systems? Who approves usage? What safeguards exist?

Second, lay out formal AI governance policies. Every AI strategy needs a board-okayed framework that covers risk assessment, human oversight, accountability, audit mechanisms and disclosure standards.

Third, review D&O policies with forensic attention. This must not be taken casually and needs close legal scrutiny. Boards should specifically examine AI exclusions, trigger events and ambiguous wording that insurers may later interpret creatively to their advantage.

Fourth, regulate employee AI usage before it gets a chance to evolve into a compliance horror story. Employees are already uploading confidential company documents into ChatGPT, mixing personal and professional AI accounts and allowing sensitive data to wander into servers located halfway across the planet. In many firms, shadow AI adoption is happening faster than official diffusion. This could be dangerous.

Finally, boards must exercise caution while using AI tools internally. Directors are found to increasingly rely on AI for summarizing reports, generating insights and reviewing disclosures. It’s convenient, no doubt, but it could return to haunt them. A flawed AI-generated summary that leads to bad oversight judgement will not magically take the blame in court.

What’s at stake is corporate governance. Many boards still approach AI as a branding exercise rather than an operational transformation that entails fiduciary consequences. Although AI tools do not ‘care’ about outcomes, regulators and stakeholders do. And insurers would rather minimize grey-zone risk coverage than take on the burden.

In India’s compliance ecosystem, confusion often travels faster than regulation. Unless care is taken, we could find litigation moving even faster. Boards must learn a harsh new rule of governance: if AI begins to run parts of the company, liability could come running right behind it.

The authors are, respectively, co-founder of the Medici Institute for Innovation.X: @MuneerMuh; and a global board advisor, coach and publisher.