What companies cut ties with Immigration and Customs Enforcement in 2018?
Executive summary
In 2018 the most widely documented corporate break from Immigration and Customs Enforcement was consulting giant McKinsey & Company, which told The New York Times it would end its work for ICE amid employee and alumni unrest in July of that year [1]. At the same time some local governments moved to sever ties with ICE-contracted private-prison operators — notably Williamson County’s June 2018 vote to end an IGSA with ICE and CoreCivic — although ICE later renewed a direct contract with CoreCivic [2].
1. McKinsey’s exit — the clearest corporate cutoff in 2018
The clearest, repeatedly reported example from 2018 is McKinsey: after internal and alumni pressure and public scrutiny the firm told The New York Times in July 2018 that it would end its work for ICE, a move that became the touchstone for employee-driven campaigns pressuring consultancies and tech firms to withdraw from immigration-enforcement work [1].
2. Local governments hitting private contractors — CoreCivic and the Williamson County case
Beyond private-sector announcements, local elected officials acted: Williamson County commissioners voted in June 2018 to end their IGSA — the intergovernmental service agreement — with ICE and CoreCivic after years of protest, a decision that temporarily cut the county’s contractual pathway to hosting ICE detainees until ICE later renewed directly with CoreCivic [2].
3. A broader wave of pressure — many companies faced calls to cut ties, but few announced formal 2018 exits
2018 also saw mass employee letters and public campaigns targeting large firms — Amazon, Microsoft, Deloitte and others — urging them to sever or restrict work for ICE and related agencies; these campaigns produced varied outcomes over time, but most of the reporting indicates protest and demands rather than immediate, blanket terminations by those corporations in 2018 [3] [4].
4. Tech-sector flashpoints: Palantir, Microsoft, Thomson Reuters and the #NoTechForICE movement
Tech vendors drew particular ire: activists pressured Palantir over a multimillion-dollar ICE contract, and campaigns like #NoTechForICE pushed Reuters and RELX to drop ties, while reporting showed Hewlett Packard Enterprise, Thomson Reuters, Microsoft, Motorola Solutions and Palantir had active contracts with ICE around that period — but public pressure produced a spectrum of responses, not a uniform pattern of 2018 terminations [5] [6] [4].
5. What counts as “cutting ties”: nuance matters
Claims that companies “cut ties” with ICE can mean different things: ending a specific contract, refusing new work, public policy changes limiting cooperation, stopping advertising or recruitment placement, or a local government terminating an IGSA with a private operator. Reporting shows McKinsey’s public withdrawal in mid‑2018 and Williamson County’s IGSA vote as concrete actions that year; many other actions cited in available sources are protests, pressure campaigns, or later-terminations and therefore cannot be confidently labeled as corporate exits in 2018 alone [1] [2] [6].
6. Alternative interpretations and reporting limits
Some sources portray additional companies as having “distanced” themselves from ICE across the broader movement, and others document later-year or later-decade contract endings — for example, airlines or ad-buy changes noted in activist reports — but the supplied reporting does not provide firm, dated evidence that those were formal corporate terminations of ICE contracts specifically in 2018, so this analysis limits affirmative claims to what the cited coverage supports [7] [8].
7. Bottom line for 2018
Based on the reporting provided, McKinsey & Company is the primary corporate example that publicly ended work for ICE in 2018, and Williamson County’s June 2018 vote ended a local IGSA with CoreCivic (which functionally cut that tie at the county level), while many other firms faced pressure and scrutiny that year without uniformly documented, formal contract terminations in 2018 itself [1] [2] [4].