How to spot an AI Deepfake
Source: YouTube · a16z · published Feb 28, 2025 · 22:31
AI models like DeepSeek are enabling increasingly sophisticated social engineering attacks, with nearly 90% of security breaches still stemming from human vulnerabilities 0:17.
Key Takeaways:
• Social engineering attacks have increased 4x since ChatGPT's emergence, with over 100,000 deepfake attacks reported in the US during 2024 2:13.
• Individuals should delete voicemail greetings to prevent voice cloning, as attackers now need only seconds of audio to create convincing deepfakes 4:51.
• Organizations must move beyond compliance-based security training to personalized, engaging education that simulates real AI-powered attacks 10:03.
• Security concerns are escalating from financial damage to potential physical harm, with threats targeting critical infrastructure like hospitals and energy systems 15:05.
The future of cybersecurity will involve an AI arms race, with defensive agents working to protect against increasingly autonomous attack agents 20:44.
Sources:
- 0:17 Discussion on how 90% of attacks come through people
- 2:13 Statistics on social engineering and deepfake attacks
- 4:51 Practical tip about deleting voicemail greetings
- 10:03 Critique of current security training approaches
- 15:05 Concerns about attacks causing physical harm
- 20:44 Discussion of AI defensive agents
Generate CPE Credits
Generate a professional CPE document from this video's transcript.
Estimated credit: 0.5 CPE hours
Estimate uses the video runtime (1 hour ≈ 1 CPE, rounded to the nearest 0.5, minimum 0.5, maximum 2.0). The final amount can be lower after review, never higher.
Topic: Cybersecurity. Commonly maps to: Security and Risk Management, Security Operations. Exact CISSP domains are assigned during generation.
Free account. One generation at a time, with a daily limit.
CPEBuddy is independent and not affiliated with or endorsed by ISC2, ISACA, or any certification body. Exports are formatted for common CPE submissions; acceptance is at your certification body's discretion.
Transcript Preview
First 800 characters of the transcript
there's still a bit of ego and hubris among the technology folks that we can secure this without securing the people that we can just secure the systems the reality is that almost 90% of attacks still come through people and if you want to talk about securing things like look I I do think there's been a lot of advancement in securing uh email as an attack factor I think that the systems are getting a lot smarter and morning some of the the systems that protect us in email and there's got to be vigilance there but the other channels are where there's uh you know just tremendous exposure so my advice to people at at companies who are in charge of security and and care about this huge new attack Factor would be to make sure that your employees are a uh extremely well educated on where these a…