A recently formed consortium of top tech companies has released its first report. The almost 200-page analysis tracks the impact of artificial intelligence on technology roles and describes what types of foundational and job-specific AI training workers will need.

The Cisco-led group includes Accenture, Eightfold, Google, IBM, Indeed, Intel, Microsoft, and SAP. Collectively, the companies have committed to train and upskill 95M people over the next decade.

The new analysis drew information from consortium members about how AI was affecting 47 top information and communication technology job roles in the U.S. and Europe, ranging from data scientist to IT support technician and business analyst. It found that 92% of those jobs are expected to undergo at least moderate transformation due to advancements in AI.

Impacts on the technical writer role might be high, for example, as the core skill of documentation can be significantly automated. Other common skills for technical writers that will be affected include scientific research, task coordination, and regulatory compliance.

To stay competitive, the report says, technical writers will need to pick up new skills, such as in automated testing and AI model development. They should focus on upskilling in project team management and other areas where AI can complement their work.

The analysis highlights the top 10 skill sets that are expected to become increasingly relevant across job roles, including AI ethics. It also identified 10 skills that may become less relevant. For example, 31% of job roles indicate a reduced need for basic software programming and language skills.

To keep pace with the changing job landscape, the report says academic institutions should prioritize investments in work-based learning, use flexible learning paths, and deepen collaboration with high schools.

“We have to act intentionally to make sure populations don’t get left behind,” said Francine Katsoudas, Cisco’s chief people, policy, and purpose officer.

Among its next steps, the consortium plans to introduce an AI workforce playbook. The companies also will explore public-private partnerships to contribute to an AI skills taxonomy that provides a common lens for skills to be defined and mapped to corresponding job roles.

Ryan Oakes, Accenture’s global health and public service industry lead, pointed to the U.S. public workforce system, which is funded by the Workforce Innovation and Opportunity Act.

In comments included in the report, Oakes said the WIOA system is “well-positioned to support this upskilling at an accelerated pace and scale in collaboration with businesses such as those leading this consortium.”