Today we celebrate: Towel Day

AI

Atlassian Cuts 10% of Workforce, AMI Labs Breaks Seed Round Record, OpenClaw Triggers Regulatory Alarm in China

Published: 2026-03-13 Updated: 2026-04-07

On March 11 Atlassian announced the reduction of around 1,600 jobs^[1], roughly 10% of its global workforce. In a memo to employees, the co-founders emphasized that the layoffs aim to “self-finance” increased investments in AI features and to expand sales to enterprise clients^[2]. The company estimates restructuring costs^[3] – including severance and office space downsizing – at $225–236 million, mostly impacting the third quarter of the fiscal year^[4]. The cuts will hit North America the hardest^[5] with about 40% of the eliminated positions^[6], followed by Australia (30%) and India (16%), with a smaller scale in Europe.

Atlassian Cuts Jobs to Fund AI

Atlassian’s decision reflects the pressure^[7] investors are putting on mature SaaS providers. Over the last 12 months, the company’s market value has dropped by over 50–60%^[8], while the market expects a convincing “AI-native” strategy and margin improvement^[9]. The company is prioritizing the development of the Rovo assistant^[10] and other agent-like features, which could slow down classic product roadmaps^[11]. This signals the diminishing relevance of specialists lacking competencies^[12] in favor of roles focused on building and integrating agents.

AMI Labs’ Record Seed Round on World Models

At the opposite end of the market, the company announced closing a seed round worth $1.03 billion^[13] at a pre‑money valuation of about $3.5 billion – the largest known seed round in Europe. AMI Labs is building so-called world models, systems learning from sensory data^[14] about the physical world rather than mainly from text. This aims to enable better planning, robot control, and applications in autonomy^[15] compared to classic language models.

The founders of AMI Labs are Yann Le. Cun^[16], serving as executive chairman, and Alexandre Le. Brun, formerly CEO of Nabla startup. Such a large concentration of capital in an alternative to LLMs strengthens Europe’s position^[17] as one of the AI development hubs alongside the USA and China. This could mean greater availability of research partnerships and grants^[18] for Polish universities and companies in robotics, logistics, and medicine.

OpenClaw Caught Between Boom and Chinese Clampdown

Simultaneously, a craze for OpenClaw^[19] – an open-source AI agent capable of independently performing computer tasks – exploded in China. The tool, created by independent developer Peter Steinberger, gained popularity^[20] thanks to easy installation and support from local clouds, but recently central institutions started formally restricting its use in banks and government agencies^[21]. CNCERT/NVDB warned that OpenClaw’s default security settings are “extremely fragile”^[22], and previously published vulnerabilities at medium and high risk levels could lead to system takeover and data leaks^[23].

The NVDB vulnerability database published a list of “6 dos and 6 don’ts”^[24] for the safe use of OpenClaw – from the principle of least privilege and avoiding unauthorized mirrors, to warnings against plugins requiring passwords. Media outlets, including Reuters and Bloomberg, report^[25] that they have received formal directives not to install OpenClaw on work computers^[26]. Paradoxically, local governments and tech hubs such as Longgang district in Shenzhen offer grants up to 10 million yuan^[27] to companies developing projects based on this agent. The OpenClaw case shows that with the rapid adoption of AI agents, the importance of requirements^[28] regarding permissions, logging, and sandboxing is growing^[29], and may set a pattern for implementing executive regulations^[30] under the EU AI Act, particularly for high-risk systems and cybersecurity.

Related posts: