AI pushes into government work
AI and automation are reshaping transportation workforces and consulting needs, while Google’s DeepMind is leaning harder into government and national-security contracts—signalling greater public-sector AI adoption. Federal officials are also calling for standardized, transparent AI evaluation and safe-path frameworks for critical infrastructure. (dcvelocity.com) (businessinsider.com) (nextgov.com)
Randstad’s Workmonitor 2026 drew on 1,752 workers, 55 employers and more than 1 million job postings to conclude AI adoption is rising while worker confidence falls — 98% of U.S. employers expect growth next year versus just 55% of talent, and 65% of employers said they’d invested in AI in the past 12 months. (dcvelocity.com) The Randstad report found stark perception gaps: 76% of employers predict at least half of entry‑level roles will disappear within five years from automation, while only 42% of workers agree, and digital confidence among workers dropped from 78% in 2025 to 64% this year. (dcvelocity.com) At an internal January town hall, DeepMind leaders including VP of global affairs Tom Lue told staff the company has a “robust process” to vet Department of Defense partnerships and said Google is “leaning more” into national‑security work, citing conversations on cybersecurity and biosecurity. (africa.businessinsider.com) DeepMind’s stance follows a 2025 update to Google’s AI principles that removed an earlier pledge against building technology “likely to cause overall harm,” a shift company leaders invoked when describing increased pursuit of government deals. (africa.businessinsider.com) The Anthropic–DoD dispute — including the Pentagon’s “supply‑chain risk” designation of Anthropic and the startup’s subsequent lawsuit — has intensified internal industry debates and prompted staff at other AI firms to sign public statements and, in some cases, press for workplace action. (africa.businessinsider.com) Federal infrastructure to standardize safe experimentation has accelerated: the GSA’s USAi evaluation suite and related guidance give agencies a cloud environment to test major models from multiple vendors, and agency officials have urged standardized, transparent model evaluation to pick appropriate models for mission needs. (usai.gov) Oversight bodies warn gaps remain for critical‑infrastructure risk assessment — a GAO review found lead agencies’ AI risk assessments need improvement — and CISA has cataloged recent federal AI actions aimed at defending critical networks and supply chains. (files.gao.gov) GSA procurement moves add near‑term buying signals: the agency’s generative AI acquisition resource guide and the March 2026 CIO directive updating AI use policies aim to speed agency adoption while codifying procurement and safety guardrails that will shape future transit‑agency RFPs and vendor evaluation criteria. (gsa.gov)
