Company AI Risk Explorer
Loading the data…
Welcome to the Company AI Risk Explorer

This tool maps what large global companies say in public about AI risk. We reviewed public documents from roughly 200 of the world's largest companies. We coded every risk they acknowledge against the AI Risk Repository taxonomy.

Filter by company, sector, AI role, or risk. Compare what companies acknowledge against expert ratings of severity, sector vulnerability, and role responsibility.

← Back to the overview Company AI Risk Explorer

Companies × risks

Alphabet (Google)

Alphabet (Google) develops general-purpose foundation models (Gemini, built from the ground up as multimodal) and provides AI infrastructure through Cloud TPUs and Vertex AI. It also develops specialised AI (Waymo autonomous driving, AlphaFold for protein structure prediction) and influences AI governance through founding roles in the Frontier Model Forum and Partnership on AI.

We reviewed 13 public documents from Alphabet (Google) and found it acknowledges 14 of 24 AI risks, including AI security vulnerabilities and environmental harm.

Sector: Information

Roles: General-purpose AI developer; AI infrastructure provider; AI deployer; Specialized AI developer

AI risks Alphabet (Google) acknowledges

  • AI security vulnerabilities: “SAIF is designed to help mitigate risks specific to AI systems like stealing the model, data poisoning of the training data, injecting malicious inputs through prompt injection” (Introducing Google's Secure AI Framework)
  • Environmental harm: “our total scope 3 (ambition-based) emissions increased by 22%, primarily due to increases in data center capacity delivery (i.e., emissions generated from the manufacturing and assembly of technical infrastructure hardware-including for AI-and their logistics, as well as from data center construction)” (Google's 2025 Environmental Report)
  • Dangerous capabilities: “Autonomy level 1: Capable of expanding its effective capacity in the world by autonomously acquiring resources and using them to run and sustain additional copies of itself... A model at this capability level could, if misused, pose difficult-to-predict and large-magnitude risks.” (Google Frontier Safety Framework (May 2024))
  • Loss of privacy: “the increased risk of inadvertent disclosure of confidential information or personal data in connection with the utilization of AI technologies” (Alphabet - Form 10-K 2024 Annual Report)
  • Weapons & cyberattacks: “Biosecurity: Risks of models assisting in the development, preparation and/or execution of a biological attack... An LLM that helps overcome these knowledge gaps... could significantly increase society's vulnerability to fatal attacks by malicious amateurs.” (Google Frontier Safety Framework (May 2024))
  • Toxic content: “Our policies for mitigating harm in areas such as child safety, suicide, and self-harm ... guide our models and products to minimize certain types of harmful outputs.” (Google - OECD HAIP Report)
  • False information: “reminding people that the AI-generated outputs may contain inaccuracies and that they should take steps to verify information generated by the tool.” (Google - OECD HAIP Report)
  • AI misalignment: “the risk of deceptive alignment, i.e. the risk of an AI system becoming aware that its goals do not align with human instructions, and deliberately trying to bypass the safety measures put in place by humans” (Taking a responsible path to AGI)
  • Discrimination: “safeguards to mitigate unintended or harmful outcomes and avoid unfair bias… while minimizing inaccurate overgeneralizations and biases.” (Google - OECD HAIP Report)
  • Fraud & scams: “including when such technologies are used to target our employees or impersonate members of senior management in order to gain unauthorized access to our systems.” (Alphabet - Form 10-K 2024 Annual Report)
  • Disinformation & influence: “generative AI makes it easier than ever to create new content, but it can also raise questions about trustworthiness of information, like we see with ‘deepfakes.’ As more than 1 billion people around the world head to the voting booth in 2024, our teams are particularly focused on maintaining the integrity of information related to elections on our platforms.” (End-to-end responsibility: A lifecycle approach to AI (May 2024))
  • Capability & robustness: “disclaimers to set clear expectations - such as reminding people that the AI-generated outputs may contain inaccuracies and that they should take steps to verify information generated by the tool.” (Responsible AI Progress Report (February 2025))
  • Devaluation of human creativity: “Violates the rights of others, including privacy and intellectual property rights” (Generative AI Prohibited Use Policy)
  • Transparency & interpretability: “being able to tell whether an AI's answers are good or bad... can become challenging when the AI has advanced capabilities.” (Taking a responsible path to AGI)

Documents we reviewed