Three stories this week put the same question on the table: who actually sets the rules for AI at work? In Washington, the White House shelved its AI safety executive order at the last minute. In Rome, Pope Leo XIV published an encyclical arguing that AI’s real risk is to human dignity, not productivity. And inside Meta, the company built a tool to watch how its own staff work, then cut 8,000 jobs. For founders and HR leaders, the thread is uncomfortable but clear. External guardrails are getting thinner, so the ones you set internally now carry more weight.

Washington Shelves Its AI Safety Executive Order

On May 21, President Trump postponed the signing of his administration’s AI safety executive order, hours before the ceremony. He told reporters he “didn’t like certain aspects of it.” He also worried it “could have been a blocker” to the US lead over China. (Source: CNBC, CNN)

What the AI safety order would have required

The draft set up a voluntary review process for new AI models before release. Federal agencies would have had up to 90 days to run a security check. Major tech, AI, and cyber CEOs were reportedly invited to the signing. Several were already en route when the event was called off. For now, the framework sits in limbo with no clear new date.

Why this matters for HR leaders and founders

Here is the practical read. Say you were waiting for a federal rulebook before tightening how your team uses AI in hiring. That rulebook just got delayed. Meanwhile, the EU AI Act and India’s DPDP rules keep moving. So a US company hiring across borders still faces real obligations abroad, even as Washington pauses at home. The gap between “legal” and “defensible” is where you now operate.

What to do: Do not treat the postponed AI safety executive order as permission to relax. Write down how your AI hiring and review tools make decisions. An audit trail protects you no matter which way federal policy lands.

Meta Built a Tool to Watch Its Own Employees, Then Cut 8,000 Jobs

Leaked details of Meta’s “Model Capability Initiative” show it captured keystrokes, clicks, mouse movements, and periodic screenshots across apps like Gmail and Slack, then used that data to train AI agents that mimic how staff work. More than 1,000 employees signed petitions calling it dystopian. Days later, on May 20, Meta cut roughly 8,000 jobs, about 10% of its workforce. (Source: TNW, CNBC)

So what does this mean for your team? Meta says the data will not feed individual performance reviews. Still, the optics are brutal. The company trained AI on its staff’s own work, then cut headcount. Its 2026 AI capex guidance runs to $115 to $135 billion. Planning to deploy AI agents into HR workflows? Tell employees what you collect and why. Do it before someone leaks it for you.

The Pope Weighs In on AI and Human Dignity

On May 25, Pope Leo XIV released his first encyclical, “Magnifica Humanitas,” on protecting human dignity in the age of AI. He signed it May 15, exactly 135 years after Leo XIII’s “Rerum Novarum” on workers’ rights. He framed the core challenge as “not technological, but anthropological.” (Source: Vatican News, NBC News)

You do not need to be Catholic for this to matter. A billion-follower institution just pushed “AI versus human dignity at work” into the mainstream. That landed the same week two big employers cut jobs and pointed at AI. Expect your own staff and candidates to start asking harder questions about how AI shapes their roles. A clear, honest answer is now part of your employer brand.

India’s Sovereign-AI Buildout Picks Up Speed

India keeps pushing to own its AI infrastructure. Larsen & Toubro has incorporated a subsidiary, Vyoma.AI, to build data centres and a sovereign cloud. Its roadmap targets above 200 MW across Mumbai, Chennai, Bengaluru, and Hyderabad. (Source: Inc42) One market report pegs the Asia-Pacific sovereign-AI infrastructure market at $9 to $14 billion in 2026.

If you hire engineers in India or run a distributed team across APAC, this trend works in your favour. More local compute and clearer data-residency options make it easier to keep workloads compliant. For teams hiring AI engineers in India, the talent and the infrastructure are now growing in the same place.

Quick Hits

• The AI skills bubble is real. A GCheck survey of 1,500 workers found 63% admit to exaggerating their AI skills. That rises to 80% among Gen Z. And 69% expect parts of their job automated within 24 months. Worth reading before your next hire, alongside our take on the AI skills gap in HR. (Source: GlobeNewswire)
• OpenAI is selling deployment, not just models. It launched a $4 billion deployment company and bought consultancy Tomoro. That brings about 150 “forward-deployed AI engineers” to embed inside enterprises. The takeaway: the hard part of AI is no longer the model, it is rolling it out. (Source: OpenAI)

What the AI Safety Order Means for Your Team

The week’s pattern is hard to miss. Public oversight slowed with the postponed AI safety executive order, while moral and market pressure picked up the slack. So the burden shifts to you: be transparent about how AI touches hiring, pay, and reviews. If you manage people across borders, Asanify’s global employer of record handles the compliance side. That frees your team to set AI policy on purpose, not in a panic.

FAQ

What was in the AI safety executive order the White House postponed?
The draft AI safety executive order would have created a voluntary review process for new AI models before release. It gave federal agencies up to 90 days for a security check. President Trump postponed the signing on May 21, 2026, saying he did not want it to slow the US lead in AI. No new signing date has been set.

Can a company legally monitor employees to train its own AI?
In many places it is legal if disclosed, but it carries real trust and morale risk. Meta’s internal monitoring program drew over 1,000 employee petition signatures. The company said the data would not be used for individual performance reviews. The safer practice is to tell employees clearly what you collect and how it will be used.

What does Pope Leo XIV’s AI encyclical mean for employers?
The encyclical “Magnifica Humanitas” argues that protecting human dignity should guide how AI is used, including at work. For employers, it raises the bar on explaining how AI affects jobs, decisions, and people. Expect more candidates and staff to ask how your AI tools treat them.

Not to be considered as tax, legal, financial or HR advice. Regulations change over time so please consult a lawyer, accountant  or Labour Law  expert for specific guidance.