Canva cuts and reshapes 22% of staff as studies point to entry-level squeeze

Canva cuts and reshapes 22% of staff

Canva announced job cuts and a sweeping overhaul that will affect more than one-fifth of its workforce.

Chief Executive Officer Melanie Perkins announced the move in a note to employees posted on the company's website. Bloomberg reported the note on 3 November. Some 1,271 people, or 22% of Canva's staff, will lose their jobs, change roles or relocate to central hubs. The report described AI as forcing a structural reckoning at one of Australia's highest-profile technology companies.

For freshers and employers, the question is which entry-level roles survive and what skills now pay. Canva's 22% figure shows restructuring is not theoretical. The studies below put numbers on the other side of it.

What the new workforce studies say

An IIT Madras white paper landed on 26 October. It argues that AI will automate tasks within jobs rather than cut entire professions.

The paper warns that automating routine junior-level tasks could weaken traditional career pathways in IT, finance and professional services. Balaraman Ravindran, head of the Centre for Responsible AI and the Wadhwani School of Data Science and AI at IIT Madras, said the paper challenges the assumption that AI inevitably takes away jobs. Gokul S Krishnan, research director at CeRAI, made a narrower point. AI taking over a task, he said, did not automatically mean a job would disappear. The outcome depends on cost, capability, regulation, safety and how organisations redesign roles.

Hiring projections from HAN Digital describe the demand side. The IT-BPM Graduate Hiring Outlook 2027 projects about 2.1 lakh fresher hires in FY2026-27, up 20% from 1.75 lakh in FY2025-26.

Pay is diverging inside those numbers. HAN Digital estimates that AI/ML engineers with AI-ready skills could command ₹8-22 lakh a year, against ₹6-12 lakh for those without them. Similar premiums show up in software engineering, full-stack development, data science, DevOps and cloud, and data analytics. A University of Phoenix white paper published on 31 October adds the worker's side: 50% of workers report learning to use AI independently, without formal employer training.

The entry-level squeeze

IIT Madras researchers identify seven pathways through which AI can transform work. They include cost-driven automation and tasks that demand human judgment.

They also describe an AI-cost geography. Affordable tools could speed automation in business process outsourcing, bookkeeping and manufacturing, while advanced work in legal research and medical imaging stays in well-resourced urban centres. HAN Digital names the roles already under pressure: manual testing, junior quality assurance, basic application support, routine data entry and traditional voice-based support.

Demand is growing at the other end. HAN Digital points to AI/ML engineers, generative AI developers and cybersecurity roles as the areas adding headcount.

Numbers in perspective

The projected 2.1 lakh fresher hires for FY2026-27 are up 20% from 1.75 lakh, but still well below the roughly 4 lakh freshers hired in FY2021-22. The recovery is real and partial at once. HAN Digital expects IT services companies to take about 1.25 lakh of the total, or 60%, with global capability centres at around 60,000.

The gap runs both ways. 60% of workers say they want more guidance in learning AI tools, and 62% of employers say employees are developing AI skills faster than their organisations can adapt. 48% of employers are concerned about retaining AI-fluent talent.

Measurement is the weak link. IIT Madras researchers say India still lacks rigorous labour-market research into which skills AI is automating or enhancing.

What happens next

IIT Madras wants cheaper, narrower tools. The researchers recommend affordable, task-specific AI tools, India-specific research on employment impacts, and structured reskilling and worker-transition pathways.

The University of Phoenix paper makes a similar point about employers. Stella Smith, associate university research chair at the university's Center for Educational and Instructional Technology Research, said rural workers may already be building valuable AI capabilities on their own, but employers need systems that help recognise, develop and apply those skills. The paper also warns that informal, uncoordinated AI adoption carries risks, including inconsistent workplace practices and data governance concerns.

Canva has not broken down the 1,271 by outcome. It has not said how many will leave the company rather than change roles or relocate.