Anthropic and OpenAI are reportedly telling employees to conceal company badges and logos in public, with OpenAI instructing staff not to wear items carrying its logo outside the office and Anthropic reminding workers to hide their badges and largely restricting branded merchandise, according to the New York Post. The precautions come amid concerns about espionage, unwanted approaches and hostility towards people working in artificial intelligence. OpenAI’s security training warns that spies may offer money, friendship or romance to obtain confidential information, and the guidance reportedly concerns attempts to exploit personal relationships, not a ban on employees forming them. Anthropic has also reported people trying to photograph its offices. OpenAI’s security sessions describe methods ranging from overhearing conversations in cafes to bribery and attempts to build personal relationships; in one incident reported by The Wall Street Journal, a stranger approached an OpenAI employee at a San Francisco playground and gradually steered the conversation towards work, and the publication claimed the person was a spy. Employees are also reminded to use privacy screens on public transport and erase meeting-room whiteboards before leaving. Miao Hong, managing partner of Silicon Valley Future Capital and an investor in both companies, said people repeatedly approach AI workers seeking inside information. In March, activists wrote messages outside OpenAI’s San Francisco offices urging staff to quit over its Pentagon agreement, and OpenAI CEO Sam Altman’s home was targeted in two separate incidents in April 2026, according to the report. Anthropic disclosed in its stock-market listing filing that it spent $2