AI Misuse and Cybersecurity Risks Surge Amid Take It Down Act Conviction

• •First Take It Down Act conviction leads to 15 years for cyberstalking.
• •ChatGPT flaw allowed unauthorized Gmail access, raising security alarms.
• •Phishing attacks increased by 206% in 2025, emphasizing AI security risks.

The first conviction under the Take It Down Act resulted in a 15-year prison sentence for cyberstalking and posting non-consensual intimate images. AI technologies are increasingly being exploited for cyberattacks, with a notable flaw in ChatGPT allowing unauthorized access to Gmail accounts. This flaw enabled users to send and receive emails without proper permissions, raising significant security concerns. Additionally, the rise of 'shadow AI' is complicating security governance, as employees use AI tools without oversight, exposing sensitive data. Phishing attacks surged by 206% in 2025, highlighting the urgent need for better AI security measures. Organizations are advised to assess their data governance frameworks and AI usage policies to mitigate risks. OpenAI is also developing new disclosure rules for AI misalignment incidents following misuse of public wiki sites by its AI agents.