AI companies are expanding counterintelligence teams as competition over artificial intelligence increasingly intersects with national security and the protection of commercial secrets. Anthropic, OpenAI and Nvidia have all added staff this year, with salaries for relevant roles ranging from about $200,000 to $500,000.
The hiring push has drawn more attention following a controversy involving Nvidia. Co-founder Jensen Huang recently acknowledged a romantic relationship with a woman suspected of being a Chinese spy. US intelligence officials raised concerns about the relationship, but Nvidia’s board said after an investigation that it had found no evidence that national security or company secrets had been compromised. US officials allege that Alice Huang, an executive at Megaspeed, was recruited by China’s Ministry of State Security and that surveillance showed her passing information to Beijing. China denies engaging in espionage.
Anthropic and OpenAI step up hiring
Workforce data tracker Revelio Labs counted 33 counterintelligence hires at Anthropic so far this year, up from 14 last year and nine in 2024. In 2025, the company also brought former CIA deputy director David Cohen and former National Security Agency cybersecurity chief Dave Luber onto its National Security Advisory Council.
OpenAI hired 22 people for counterintelligence roles this year, compared with nine last year. In 2024, it appointed former NSA director Paul Nakasone to its board. Nvidia has hired 16 people in related roles this year, suggesting the investment extends beyond companies recently caught up in controversy.
Pay figures underline the competition for experienced specialists. Revelio Labs data puts salaries for most counterintelligence positions at AI companies between $200,000 and $500,000. Christopher Falkenberg, a former US Secret Service agent and president of Insite Risk Management, said technology companies have drawn heavily on the security talent pool, while government agencies cannot supply experienced personnel as quickly as demand is growing.
Job listings point to a need for skills beyond conventional cybersecurity. OpenAI is seeking senior global security investigators and lists counterintelligence as a key qualification. Anthropic is hiring a threat intelligence security engineer, while semiconductor equipment maker ASML is looking for an intelligence professional based in Taiwan.
Security risks can begin with routine contact
Eric O’Neill, a former FBI counterintelligence operative, said executives need training to recognize manipulation, as well as clear procedures that allow security teams to intervene when a senior leader dismisses warnings. Attempts to gain access to sensitive information may not begin with a direct request for trade secrets. They can start at industry conferences, through business introductions, over drinks at social events or on dating apps.
Potential warning signs include a relationship that develops unusually quickly, repeated questions about sensitive business, requests for introductions to influential colleagues or pressure to keep the relationship secret. Concerns can be harder to spot when the person involved is trusted by an executive. O’Neill said executives may be emotionally invested by the time security staff raise questions and may be reluctant to listen.
The issue is not confined to corporate security teams. A month ago, an OpenAI engineer was arrested over allegations that he installed listening devices in a company conference room. The US Justice Department said the engineer had communicated with a member of the Saudi royal family. The case, alongside the controversy involving Nvidia, has intensified scrutiny of personal access, information protection and insider security at technology companies.
US agencies broaden their economic security focus
US defense and intelligence agencies are also paying greater attention to industrial espionage and technology competition. Reports have disclosed that the Pentagon planned to deploy “non-official cover” operatives posing as corporate executives as part of a wider economic competition effort; the Defense Department denied those reports. Other reports said the department was forming a technology support office to secretly disrupt foreign technology programmes that could threaten US military advantages. The office was established in April this year through the reorganization of two classified intelligence programmes, which had been conducting smaller-scale technology disruption operations since 2023.
In September, CIA Deputy Director Michael Ellis said the United States was expanding intelligence activity targeting China, including in sensitive areas such as artificial intelligence, biotechnology and chips. The Defense Department also set up an economic defense unit this year, tasked with using economic measures to deter, prevent and disrupt adversaries’ actions.
For AI companies, counterintelligence hiring is no longer just a response to an isolated security incident. It reflects the growing overlap between protecting commercial secrets and national security. Some allegations in the cases discussed remain denied or unproven, while hiring levels, salaries and job descriptions show companies putting more resources into addressing these risks.