While crucial, the slow, administrative, and sometimes political process of defining "responsible AI" is becoming a deterrent for pharma companies. Aditya Gherola argues that regulators must move faster to provide clear guidelines, preventing the concept from becoming a roadblock to critical innovation in drug discovery.

The vocabulary of AI safety and regulation (e.g., 'national security threats,' 'autonomy risk') is so ambiguous that a power-hungry government could easily abuse it. Any AI model that refuses government orders, such as for mass surveillance, could be labeled an 'autonomy risk' and shut down, creating a pre-built tool for despotism.

Many AI safety guardrails function like the TSA at an airport: they create the appearance of security for enterprise clients and PR but don't stop determined attackers. Seasoned adversaries can easily switch to a different model, rendering the guardrails a "futile battle" that has little to do with real-world safety.

Policymakers confront an 'evidence dilemma': act early on potential AI harms with incomplete data, risking ineffective policy, or wait for conclusive evidence, leaving society vulnerable. This tension highlights the difficulty of governing rapidly advancing technology where impacts lag behind capabilities.

Large firms prioritize protecting existing assets, leading to a "risk-first" mindset. This causes them to delay AI deployment by trying to eliminate all potential downsides—a futile effort that stalls innovation and makes them vulnerable to disruption by nimbler startups.

Governments face a difficult choice with AI regulation. Those that impose strict safety measures risk falling behind nations with a laissez-faire approach. This creates a global race condition where the fear of being outcompeted may discourage necessary safeguards, even when the risks are known.

The greatest risk to integrating AI in military systems isn't the technology itself, but the potential for one high-profile failure—a safety event or cyber breach—to trigger a massive regulatory overcorrection, pushing the entire field backward and ceding the advantage to adversaries.

Large organizations' natural 'risk-first' mindset leads them to try and reduce all potential AI-related errors to zero before implementation. Hoffman argues this is an impossible task that prevents progress, comparing it to refusing to drive a car until every conceivable road risk is eliminated.

Leaders adopt advanced AI to accelerate innovation but simultaneously stifle employees with traditional, control-oriented structures. This creates a tension where technology's potential is neutralized by a culture of permission-seeking and risk aversion. The real solution is a cultural shift towards autonomy.