Samsung Strike Threat Grows; Bank of Canada: AI Isn't Replacing Workers at Scale

Samsung's AI memory success may be sowing internal friction while central bank research finds AI-driven labor displacement is moving slower than early models projected.

📊 Reuters Exclusive: Samsung AI Boom Fueling Strike Threat and Deep Internal Divisions

Decoded: Reuters published an exclusive investigation on May 15 revealing that Samsung's surging AI chip business — driven by demand for high-bandwidth memory (HBM) used in Nvidia GPU clusters — has created deep internal divisions and a looming labor strike at the company. Samsung employees are split between chip engineers and production workers, with floor staff warning that AI output targets have worsened working conditions. South Korea's government said on May 17 it would pursue all available options to prevent a Samsung strike. Samsung is the world's largest DRAM producer and a primary supplier of HBM3E memory used in Nvidia's Blackwell and Hopper GPU architecture; it competes with SK Hynix for the highest-margin HBM contracts. (Reuters, May 15 and May 17, 2026)

Why it matters: A Samsung production strike would introduce the first major supply chain disruption to HBM availability since AI chip demand accelerated in 2023. Samsung supplies HBM to Nvidia and competes with SK Hynix, which currently holds the largest share of high-margin HBM3E contracts. A strike scenario benefits Micron (MU) — the primary U.S. HBM supplier — which would be positioned to absorb displaced orders. For Nvidia (NVDA) investors, Samsung labor risk is the kind of low-probability, high-impact supply chain variable that is difficult to price until it materializes — but historically creates short-term volatility when it does.

🔬 Bank of Canada: AI Is Not Replacing Workers on a Large Scale — Yet

Decoded: The Bank of Canada published research on May 17 finding that AI is not replacing workers on a large scale so far. The research found that while AI adoption is accelerating among knowledge workers, aggregate employment displacement attributable to AI remains statistically modest at the economy-wide level — consistent with prior technology cycles where adoption leads displacement by several years. Bank of Canada researchers noted that current AI tools are augmenting rather than replacing most work roles, and that observable productivity gains are not yet translating into headcount reductions across the private sector at scale. The findings align with U.S. Bureau of Labor Statistics data and OECD research on near-term labor market disruption. (Reuters, May 17, 2026)

Why it matters: Central bank labor research directly informs policymakers calibrating AI regulation, rate-setting models, and government AI investment frameworks. If AI displacement is slower than feared, regulators face less pressure to impose emergency safeguards — a structural positive for AI developers and enterprise deployers in regulated verticals. For investors modeling the AI diffusion curve, a central bank finding that AI is augmenting rather than replacing suggests the productivity adoption cycle is still in its early-to-mid phase — a signal that enterprise AI spending may sustain longer than bearish models assume.

Stay decoded. See you tomorrow.

— The Get AI Decoded Team