AI-driven malicious bots now dominate the internet for the first time, accounting for 51% of all web traffic while launching increasingly sophisticated attacks on APIs and user accounts with 40% more account takeovers than last year.
DeepSeek, a Chinese AI platform accused of data theft, spreading propaganda, and violating U.S. export controls by covertly funneling American user data to the CCP and training on material unlawfully obtained from U.S. AI models, has drawn scrutiny from Congress over Nvidia chip sales.
In a massive ad fraud crackdown, Google suspended an unprecedented 39.2 million advertiser accounts in 2024, tripling the previous year's figure, by leveraging AI and policy updates to combat business impersonation, deepfakes, and illegitimate payment details.
OpenAI plans to verify organizations before granting access to advanced AI models, prioritizing safety by mitigating potential misuse, intellectual property theft, and unintended consequences while enabling responsible development.
As AI agents proliferate, traditional authentication methods face challenges, prompting the emergence of innovative approaches like dynamic credential issuance, identity-aware proxies, and ephemeral compute with baked-in identity to ensure frictionless yet secure access.
A sweeping coalition of Nobel laureates, labor unions, faith leaders, and entertainers launches the Pro-Human AI Declaration, demanding strict AI safeguards including bans on unsafe superintelligence development, mandatory off-switches, child protections, and independent oversight to keep humans firmly in control of increasingly powerful AI systems.
DeepMind's 145-page paper warns of potential 'severe harm' from advanced AI by 2030, proposing safeguards like access controls and AI interpretability, though some experts dispute the premises.
Chinese companies employ various tactics, including smuggling, exploiting loopholes, leveraging subsidiaries, utilizing cloud services, and innovating with alternative chips, to circumvent U.S. restrictions on advanced semiconductor technology.
An AI image generator's unsecured database exposed over 95,000 records, including disturbing prompts and explicit images, some potentially illegal, such as de-aged celebrities made to look like children, leading the company to delete its websites after being contacted.
As AI agents gain autonomy across applications, offering convenience, it's crucial to strike a balance by maintaining human oversight and control to mitigate privacy, safety, and security risks and ensure AI serves human interests.