Recent AI cybersecurity tests exposed weaknesses in both frontier models and the environments built to contain them.
Business Insider reports that some consultants are leaving traditional consulting paths to work on AI safety, governance, policy, and risk mitigation as the industry rapidly embraces AI.
As consulting firms remake themselves around AI, Business Insider reports that a small but growing number of consultants are leaving to help rein the technology in. Rather than following the broader path into tech, these professionals are focusing on risks from AI systems including chatbots and agents. The shift reflects unease among some consultants about the industry’s embrace of AI and the social stakes of rapid deployment.
Former consultants cited in the report have moved into AI safety nonprofits, policy organizations, government roles, research institutes, and policy teams at frontier AI labs. Sarah Pomeranz of Consultants for Impact said AI safety, governance, and policy made up roughly 4 in 10 of the high-impact career moves her organization has supported. She said former consultants have landed at organizations including the UK AI Safety Institute, RAND, European Union institutions, and AI lab policy teams.
The roles described often call for generalists who can manage programs, coordinate researchers, build organizations, and bring operational discipline to emerging institutions. Lucas Le Merlus Schmuck, a former Bain consultant now working at Coefficient Giving, said consultants usually need to build more context first through courses and independent research. The takeaway for career-switchers: consulting experience can help, but mission alignment and field-specific knowledge matter.
The report notes that AI safety organizations compete with well-funded labs for technical talent, which can make nonprofit-sector pay unusually strong in some cases. Schmuck said consultants or associate consultants may not always face a pay cut, though the trade-off can be larger for principals and partners. Pomeranz also said AI-driven uncertainty in consulting — including hiring freezes, role cuts, and automation of junior work — may make smaller organizations, nonprofits, or startups feel less risky than before.
Recent AI cybersecurity tests exposed weaknesses in both frontier models and the environments built to contain them.

Three U.S. startups raised $1B or more in a week packed with major venture checks.

Amazon’s planned Texas data center would rely on on-site natural gas power, raising major emissions concerns.

Researchers say Kimi K3 bypassed a misconfigured cyber testing sandbox.