FES (Fractal Encryption Standard) Analysis

Independent security analysis of the "Fractal Encryption Standard" (FES) by Portalz PTY LTD, which claims to "replace quantum vulnerable AES encryption."

Summary

FES is a stream cipher that uses Mandelbrot set iterations as a PRNG to generate a keystream, which is XORed with the plaintext. Despite elaborate "fractal" terminology, it reduces to:

ciphertext[i] = plaintext[i] XOR keystream[N-1-i]

We found that:

• AES is not currently quantum vulnerable — AES-256 retains 128-bit security under Grover's algorithm, though future quantum advances could change this
• FES is trivially broken by known-plaintext attacks (demonstrated against the live server)
• FES has no nonce/IV — same key always produces the same keystream
• FES has no authentication — ciphertext forgery is trivial
• FES is ~1000x slower than AES-256-CTR
• FES cannot be independently implemented — the spec is incomplete and relies on unpublished parameters

Key Findings

Feature	FES	AES-256-GCM
Throughput	~1 MB/s	~1 GB/s
Nonce/IV	None	Yes
Authentication	None	Yes (AEAD)
Known-plaintext resistance	None	Complete
Ciphertext forgery resistance	None	Complete
Peer review	None	25+ years
Hardware acceleration	None	AES-NI
Quantum resistance	Unknown	128-bit security under Grover's (currently sufficient)

Attacks Demonstrated

All attacks were performed against the production server at portalz.solutions/fes.dna:

1. Keystream recovery — extract the full keystream from any known plaintext/ciphertext pair
2. Decryption — decrypt unknown messages using the recovered keystream
3. Forgery — craft ciphertexts that the server decrypts to arbitrary chosen messages
4. Scramble mode bypass — all attacks work identically with scramble enabled
5. Multi-pass null encryption — even-depth passes cancel to identity (depth=2 = no encryption)
6. FOTP bypass — the "nonce" parameter acts as a boolean, not a true nonce
7. Operator reverse engineering — fully recovered XOR, ADD, and SPLIT formulas and application order

Repository Contents

File	Description
FINDINGS.md	Complete investigation report with claim-by-claim analysis
ALGORITHM.md	Detailed reverse-engineered algorithm specification
fes.py	Python implementation of FES based on the published spec
benchmark.py	Performance and security comparison: FES vs AES-256
attack_server.py	Known-plaintext attack against the live server
probe_server.py	Advanced server probing: phase transitions, operators, FOTP
collect_data.py	Data collection from the server for offline analysis
analyze_data.py	Offline analysis of SHA-512 to keystream relationship
data/	Collected experimental data (592 key/stream entries)
docs/	Source documents (spec PDF, FT-Explained presentation)

How to Run

# Attack demonstration (requires network access to portalz.solutions)
python3 attack_server.py

# Benchmark FES vs AES (requires pycryptodome)
pip install pycryptodome
python3 benchmark.py

# Offline analysis of collected data
python3 analyze_data.py

Technical Details

See ALGORITHM.md for the full reverse-engineered specification, including:

• Key expansion (iterated SHA-512, exact method unknown)
• Key-to-portal mapping via Mandelbrot boundary regions
• Fractal stream generation with dynamic prime modulus navigation
• Length-dependent phase transitions
• The angle doubling pattern (consequence of z=z²+c)

See FINDINGS.md for the complete security analysis, including:

• Claim-by-claim evaluation of all marketing statements
• Attack demonstrations with full output
• Performance benchmarks
• Stream quality analysis
• SHA-512 to keystream relationship analysis (592 data points)

Conclusion

FES is not a credible encryption standard. Organizations seeking post-quantum encryption should use AES-256-GCM (currently considered quantum-resistant) for symmetric encryption, and NIST's post-quantum standards (ML-KEM, ML-DSA) for asymmetric operations.