<div id="readability-page-1"> <time datetime="2026-05-13T06:00:00+01:00"> Wednesday 13 May 2026 6:00 am </time> | Updated: <time datetime="2026-05-12T15:02:56+01:00"> Tuesday 12 May 2026 3:02 pm </time> <div> Add as a preferred source on Google </div> <figure> <div> </div> <img width="742" height="495" src="https://imageproxy.pbkrs.com/https://www.cityam.com/wp-content/uploads/2026/01/GenAI-e1767955660131.jpg/query-dz03NDI?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" original-src="https://imageproxy.pbkrs.com/https://www.cityam.com/wp-content/uploads/2026/01/GenAI-e1767955660131.jpg/query-dz03NDI"/> <figcaption>Banks are rushing into AI.</figcaption> </figure> Atlassian claims businesses are still struggling to turn the AI boom into real company-wide productivity gains, as the software giant pushes deeper into workplace automation with a new generation of AI agents.According to the software giant’s chief AI officer, Tamar Yehoshua, most companies have moved beyond simply experimenting with AI tools, yet many were still failing to embed them into the day-to-day running of their organisations.“Individual productivity is increasing, but not the overall productivity of the organisation,” Yehoshua told City AM. “That’s the big gap right now.”The comments come as firms globally face mounting pressure to justify huge spending on AI software, amid growing scrutiny over whether the technology is genuinely delivering measurable returns.Atlassian’s products, including Jira and Confluence, are used by 85 per cent of Fortune 500 companies, as the firm bets that the next phase of AI adoption will be driven by agents.The Australian tech company recently announced a raft of new AI products centred around its workplace data layer that maps how companies operate across projects, documents, meetings, tickets and employees.The system now contains more than 154bn connections spanning Atlassian products and external platforms like Microsoft, Google and Figma, according to the company. Yehoshua described it as “a map of how organisations work”.“We understand the connections between people, projects, documents, tickets and workflows,” she said.Atlassian also recently announced it would open the teamwork graph to external AI agents and copilots through new developer tools and integrations with Microsoft Teams and Copilot.<h2 id="h-the-rise-of-workplace-agents">The rise of workplace agents</h2>Silicon Valley is shifting away from AI assistants that simply answer questions and towards systems that can actively perform work.“The agents are more powerful because they take action,” Yehoshua said. “You’re not doing just a one-off query anymore. You’re doing something that has an action associated with it”.She described internal use cases where AI agents analyse customer meeting transcripts, identify recurring complaints, create Jira tickets, assign tasks to engineers and draft follow-up emails automatically.“Nobody writes the first draft of documents anymore” she said. “Product managers don’t write help documents anymore. They say: ‘Read all the design docs, look at the code and write a draft of this.'”Mercedes-Benz has already used Atlassian’s agents to process bug reports from vehicle test fleets, improving report quality by 90 per cent while cutting duplicate detection time by 85 per cent, according to the company.The push comes amid a wider debate over whether AI will reduce workloads or simply increase expectations on workers.Businesses have been increasingly warned that productivity gains from AI could end up fuelling more work rather than less, a modern version of the ‘Jevons paradox’, where efficiency gains simply lead to higher output demands.Yehoshua argued AI should instead free employees to focus on more strategic work: “A product manager used to have to take notes in meetings, send out action items, file bugs and write documentation”.“They don’t have to do those things anymore”, she added. “They can spend more time with customers and focus on what we should be building.”The company’s AI expansion also comes amid growing scrutiny over how enterprise software groups use customer data to train AI systems.Atlassian recently disclosed plans to collect certain metadata and in-app data from customer cloud products unless enterprise customers opt into stricter controls.Yehoshua said Atlassian does not build its own large language models and instead routes tasks through an AI gateway, capable of connecting to dozens of third-party models depending on cost, performance and task requirements.Despite mounting concerns over job displacement, Yehoshua said the bigger challenge for businesses was adapting culturally to AI rather than dealing with a technical skills shortage.“People have done their jobs in a certain way for decades,” she told City AM. “The people who embrace it and say ‘this is really cool, I want to learn about it’ – they succeed.”. </div>

TEAM

北美科技軟件股指數 ETF - iShares

標普軟件與服務 ETF - SPDR

Atlassian 的人工智能負責人 Tamar Yehoshua 強調，儘管公司正在投資於人工智能，但他們在實現顯著的生產力提升方面仍然面臨困難。儘管個體生產力有所改善，但整體組織效率依然較低。Atlassian 正在推出新的人工智能產品，以增強工作場所的自動化，旨在連接各種運營元素。該公司強調向執行任務的人工智能代理的轉變，而不僅僅是回答查詢。Yehoshua 認為，人工智能應該使員工能夠專注於戰略性工作，儘管關於工作崗位流失和文化適應人工智能的擔憂依然存在

Atlassian 的 AI 負責人表示：企業仍未能使 AI“真正高效”