The burgeoning enthusiasm surrounding artificial intelligence is bordering on speculative mania, according to Meredith Whittaker, CEO of encrypted messaging app Signal. In an interview with Der Spiegel, Whittaker expressed concerns that the current industry valuation and development trajectory constitute a bubble, characterizing the prevailing atmosphere as “more show than substance.
Whittaker, also co-founder of the AI Now Institute at New York University, highlighted what she termed “circular investments” fueling the hype. She specifically pointed to the proposed $100 billion investment by Nvidia into OpenAI, the developer of ChatGPT, coupled with OpenAI’s subsequent reliance on Nvidia’s chips. This cycle, she argues, echoes the dynamics of the dot-com bubble, creating a self-perpetuating cycle of inflated expectations and financial commitments. Signal, she affirmed, has no current plans to develop its own AI technology.
Beyond the financial aspects, Whittaker voiced serious anxieties regarding the accelerating development of AI agents and their automation capabilities. She warns that these advancements pose a significant “existential threat” to both cybersecurity and individual privacy. The requirement for these AI assistants to access vast quantities of personal data across numerous apps to be genuinely useful is, she states, “extremely worrying.
The proliferation of AI agents, Whittaker argues, dramatically expands the attack surface for malicious actors, potentially undermining the security of systems. She explicitly warned that the emergence of these advanced AI programs could make it increasingly difficult, if not impossible, for Signal to maintain its commitment to app-level security. This situation raises fundamental questions about the unchecked expansion of AI and its potential to compromise the secure digital spaces individuals rely upon.
