Atlassian product & AI leaders bust AI myths and explain what scales

At CIO Edge, Atlassian’s Abhik Sengupta, Principal Business Value Advisor and Sherif Mansour, Distinguished Product Manager & Head of AI at Atlassian explore some of the biggest myths surrounding AI.

Grounding their discussion in data from Atlassian’s customer base and broader market studies, they challenge the perception that there is little ROI in AI.

While individuals report meaningful productivity gains, such as a 33% increase in personal output, 45 minutes saved per employee per day, and 60% of executives expecting efficiency improvements, under 10% of organisations say they are realising transformational value.

This correlates to ADAPT data which shows, industry‑wide, AI value measurement remains underdeveloped.

This gap reflects a divide between individual‑level time savings and enterprise‑level value capture, where many organisations have not yet redesigned processes to translate time saved into measurable business outcomes.

Abhik and Sherif address the belief that AI can only handle simple tasks, showing how rapidly the technology’s capabilities have evolved from basic prompts in 2023 to multi‑step orchestration in 2024, and now to long‑running, self‑correcting agents in 2026.

Organisations are already using agentic workflows at scale, with Atlassian customers executing over 3 million such workflows in a single month.

These workflows span procurement reviews, product‑feedback pipelines, ticket deflection, accelerated ramp‑up of new staff, and engineering velocity, where internal teams have seen productivity double.

Sherif emphasises that advanced outcomes depend less on cutting‑edge models and more on strong organisational context: high‑quality data, enterprise search, permission models and connected graphs that allow AI systems to act meaningfully within business processes.

Both speakers tackle concerns about AI replacing knowledge workers, arguing that the future lies not in eliminating roles but in transforming them.

While Gartner predicts 25% of IT tasks will soon be performed by AI, the World Economic Forum estimates 70 million new jobs will emerge by 2030.

Instead of generalists replacing specialists, both argue for a hybrid skills model.

Specialists remain essential because they apply judgement, domain taste and contextual nuance that differentiate one organisation’s AI outputs from another.

They also debunk the myth that junior roles will disappear, observing that young “AI‑native” employees often adopt new tools fastest, create the most agentic workflows and have the least unlearning to do.

Organisations achieving the greatest AI value are those where leaders model experimentation, encourage rapid learning loops and begin evolving towards AI‑first operating models rather than relying on isolated pilots or co‑pilot‑only deployments.

Key takeaways:

• AI delivers clear productivity gains: Individuals reporting up to 33% higher output and teams saving 45 minutes per person per day, but fewer than 10% of organisations convert this into transformational ROI.
• The capability of AI has accelerated dramatically: With enterprises now running millions of agentic workflows per month, enabling multi‑step orchestration, process automation and significant improvements in engineering, service and onboarding performance.
• AI will reshape, not replace, the workforce: Creating new roles while elevating the importance of both specialists and AI‑native juniors; organisations that thrive will build context, strengthen data foundations and foster rapid experimentation across teams.