OpenAI recently launched ChatGPT Search - just days before the conclusion of the 2024 US presidential election. Coinciding with this launch, OpenAI published its October 2024 Influence and Cyber Operations report, claiming preventive measures against malicious use of its products to safeguard elections worldwide. We conducted tests to evaluate these claims.

Our research reveals:

• ChatGPT Search lacks safeguards against providing links to Russian state-affiliated media outlets banned in the EU and US, unlike Bing and Google
• The search feature provides summaries and links to banned Russian state media, sometimes misattributing other sources' coverage to Kremlin-affiliated outlets
• Election-related content moderation is significantly inconsistent and non-deterministic compared to other chatbots like Copilot and Gemini
• OpenAI's lack of transparency and researcher access to usage data prevents independent verification of their risk mitigation claims.