Anthropic

Google DeepMind

OpenAI

Microsoft

xAI

Final Score:

61

42

46

25

25

Acknowledgement of AI Moral Status:

5

2

2

2

1

Transparency on AI Capabilities and Limitations:

4

3

5

2

2

Employee and Stakeholder Awareness and Training:

5

4

6

2

3

AI Rights and Protections:

5

3

2

3

1

Accountability for AI Systems:

6

6

5

3

3

Commitment to Safety in AI Development:

9

6

4

3

3

Robustness and Resilience to Adversarial Attacks:

6

5

4

5

3

Transparency and Explainability:

7

3

5

2

2

Mitigation of Bias and Unintended Outcomes:

6

5

6

1

3

Collaboration with External Experts and Safety Research:

8

5

7

2

4